A Critical

Introduction to the
Metaphysics of
Modality

9781472524263_txt_print.indd 1 01/12/2015 09:44

 BLOOMSBURY CRITICAL INTRODUCTIONS TO
CONTEMPORARY METAPHYSICS

Editorial Board:
Bill Brewer, King’s College London, UK; Albert Casullo, University of Nebraska-
Lincoln, USA; Thomas M. Crisp, Biola University, USA; Peter Simons, Trinity
College Dublin, Ireland; Tuomas Tahko, University of Helsinki, Finland

Bloomsbury Critical Introductions to Contemporary Metaphysics introduces

and advances the central topics within one of the most dynamic areas of
contemporary philosophy.

Each critical introduction provides a comprehensive survey to an important

metaphysical subject or question. Covering the historical, methodological
and practical contexts, it identifies and explores the major approaches,
theories and debates. Capturing the changes to the ways the discipline is
being studied, the emphasis placed on the historical background allows
connections to be made between contemporary issues and the wider
history of modern philosophy.

Designed for use on contemporary metaphysics courses, these introductions

are defined by clarity of argument and equipped with features to facilitate
and encourage further study. The result is a series of essential introductions
for upper-level undergraduates and postgraduates wishing to stay informed
of the issues and arguments shaping twenty-first century metaphysics.

TITLES IN THE SERIES INCLUDE:

A Critical Introduction to Causal Powers and Dispositions, Ruth Porter Groff

A Critical Introduction to Fictionalism, Frederick Kroon, Jonathan McKeown-
Green and Stuart Brock
A Critical Introduction to Properties, Sophie R. Allen
A Critical Introduction to the Metaphysics of Time, Benjamin L. Curtis and
Jon Robson

9781472524263_txt_print.indd 2 01/12/2015 09:44

 A Critical
Introduction to
the Metaphysics of
Modality
ANDREA BORGHINI

Bloomsbury Academic
An imprint of Bloomsbury Publishing Plc

9781472524263_txt_print.indd 3 01/12/2015 09:44

 Bloomsbury Academic
An imprint of Bloomsbury Publishing Plc

50 Bedford Square 1385 Broadway

London New York
WC1B 3DP NY 10018
UK USA

www.bloomsbury.com

BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc

First published 2016

© Andrea Borghini, 2016

Andrea Borghini has asserted his right under the Copyright, Designs and Patents Act, 1988,
to be identified as the Author of this work.

All rights reserved. No part of this publication may be reproduced or transmitted in any
form or by any means, electronic or mechanical, including photocopying, recording, or
any information storage or retrieval system, without prior permission in writing from the
publishers.

No responsibility for loss caused to any individual or organization acting on or refraining

from action as a result of the material in this publication can be accepted by Bloomsbury or
the author.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

ISBN: HB: 978-1-4725-2426-3

PB: 978-1-4725-2526-0
ePDF: 978-1-4725-3390-6
ePub: 978-1-4725-2194-1

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress.

Series: Bloomsbury Critical Introductions to Contemporary Metaphysics

Cover image © Mmdi/Getty Images

Cover designer: Catherine Wood

Typeset by Fakenham Prepress Solutions, Fakenham, Norfolk NR21 8NN

9781472524263_txt_print.indd 4 01/12/2015 09:44

 Contents

Acknowledgments vii

Introduction: Framing the debate 1

1 The metaphysics of modality: A historical overview 19

2 Modal skepticism and modal expressivism 49
3 Modalism 73
4 Modal realism 91
5 Ersatzism 115
6 Modal fictionalism and modal agnosticism 139
7 The new modal actualism 157
8 Necessities, necessary existents, and their bounds 175

Epilogue: Modal talk and the analyses of modality 195

Bibliography 199
Index of Names 219
Index of Terms 221

9781472524263_txt_print.indd 5 01/12/2015 09:44

9781472524263_txt_print.indd 6 01/12/2015 09:44
 Acknowledgments

W riting of this book took place during a sabbatical leave granted by the
College of the Holy Cross, and it was possible thanks also to a Publication
and Research grant from the College of the Holy Cross, through which I could
secure the impeccable assistance of Natalia Iacobelli in the early stages of
the writing process. The book vastly develops, expands, and refines an earlier
volume of mine, published in Italian by Carocci Editore, in 2009—Che cos’è
la possibilità (Possibility: What Is It? (literal translation)). I thank the editorial
staff of Carocci, and in particular Gianluca Mori, for allowing me to rework the
ideas contained in that volume. Along the way I have benefited immensely
from discussions of the themes considered here with numerous people. I
wish to thank them all; among their great number is Adriano Angelucci, Sergio
Bernini, Claudio Calosi, Andrea Cantini, Massimiliano Carrara, Elena Casetta,
Annalisa Coliva, John Collins, John Divers, Maurizio Ferraris, Salvatore Florio,
Laura Franklin-Hall, Pierluigi Graziani, Chris Haufe, Andrea Iacona, Philipp
Keller, Vittorio Morato, Luca Morena, Luca Moretti, Stephen Mumford, Marco
Nathan, Andrea Sauchelli, Giacomo Sillari, Matthew Slater, Giuliano Torrengo,
Vera Tripodi, Sebastian Watzl, and Neil Williams. Thanks to Massimo Mugnai,
who first acquainted me with the history of modality and contemporary
theories of possibility, and who punctually followed my early research on the
topic; and to Tommaso Tempesti for suggesting that I choose the topic for
my Laurea thesis. I am much indebted to my colleague Carolyn Richardson,
who provided superlative editorial assistance while reviewing the manuscript.
I owe special gratitude to Achille Varzi for the continuous incitement and
encouragement provided over the years. Finally, without Ave, Guido, Lisa, and
Elena, not only the book you hold in your hands, but much more would not
have been realized (of me and what is mine).
I dedicate this book to my grandfather Dario, his brother Guido, and their
parents, Angiola and Angiolo: that our reflection on what can be will prompt
us to teach more vividly that violence against innocent civilians cannot be.

9781472524263_txt_print.indd 7 01/12/2015 09:44

9781472524263_txt_print.indd 8 01/12/2015 09:44
 Introduction:
Framing the debate

T he introduction presents the main framework for the book and its structure.
It starts by explaining the problem of possibility according to three main
perspectives: semantic, epistemic, and metaphysical (§0.1.). The perspectives are
closely entrenched and, for each modal theory, one of them shall be employed
as a privileged point of departure (§0.2.). The first tool required for a discussion
of modality is an appropriate vocabulary: some basic concepts and terms are
presented (§0.3.). Next on the menu is an illustration of the key methodological
and structural choices that characterize the book (§0.4.), followed by a map of
the theories of possibility, through which the content of the volume is illustrated
(§0.5.). Finally, the possibility of impossible scenarios is brought up, and its
philosophical interest is discussed (§0.6).

But if there is a sense of reality, and no one will doubt that it has its justification
for existing, then there must also be something we can call a sense of
possibility. Whoever has it does not say, for instance: Here this or that has
happened, will happen, must happen; but he invents: Here this or that might,
could, or ought to happen. If he is told that something is the way it is, he
will think: Well, it could probably just as well be otherwise. So the sense of
possibility could be defined outright as the ability to conceive of everything there
might be just as well, and to attach no more importance to what is than to what
is not.
R. MUSIL, THE MAN WITHOUT QUALITIES, TRANS. B. PIKE AND S. WILKINS,
KNOPF, 1995: I, 10–11

9781472524263_txt_print.indd 1 01/12/2015 09:44

 2 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Why read a book on the metaphysics of possibility? At first one could argue
that it is an abstruse topic reserved only for those with a strong philosophical
disposition. Practical individuals are interested in who they are and what they
do, what there is and what happens: “I am Italian”; “I have US $30,000 in the
bank”; “Napoleon was defeated at Waterloo.”
This is perhaps one of the most misleading observations that can be made
regarding the topic, and it is so for at least three reasons.

Reason number one: If we take a moment to reflect, we find

possibility nestled in the most familiar of things: even in things that
at first glance concern what there is and what happens. Being Italian
means, among other things, that I am able to freely travel throughout
the European community; having US $30,000 in the bank means,
among other things, that I am able to buy certain goods; Napoleon’s
defeat at Waterloo is dramatic and meaningful precisely because
Napoleon could have won. In that which is, and in that which
happens, there are the seeds of what could be and will be, of what
could happen and will happen. (As we will have a chance to see,
some even claim that that which is and that which happens are what
they are precisely because they contain these seeds.)

Reason number two: If something is or happens, that means that

it was possible for it to be or happen. Actuality, therefore, is but one
of the faces of possibility. And, as we shall see in the next chapter,
this has been a founding principle in theorizing modality since
Aristotle, summarized in the Latin dictum: ab esse ad posse valet
consequentia (“from the fact that something happens, it follows that
it is possible”).

Reason number three: While the majority of possible things are not
attemptable, possibilities play a primary role in our lives. We might
consider the relevance of possibility in explaining our frames of
mind and our judgments: we do not need to watch an atomic bomb
explode in order to be scared of one, and we do not need to see a
homicide in order to accuse someone of attempted homicide.

So, why read a book on possibility? Because we are immersed in it, much
more than it might appear at first glance. In the remainder of this intro-
duction, we shall first identify the main philosophical questions arising for
any theorizing about possibility, thus explaining the problem of possibility
according to three main perspectives: semantic, epistemic, and metaphysical
(§0.1.). The perspectives are closely entrenched and, for each modal theory,
one of them shall be employed as a privileged point of departure (§0.2.). The

9781472524263_txt_print.indd 2 01/12/2015 09:44

 INTRODUCTION: FRAMING THE DEBATE 3

first tool required for a discussion of modality is an appropriate vocabulary:

some basic concepts and terms are presented (§0.3.). Next on the menu
will be an illustration of the key methodological and structural choices that
characterize the book (§0.4.), followed by a map of the theories of possibility,
through which the content of the volume is illustrated (§0.5.). Finally, we shall
briefly consider the possibility of impossible scenarios (§0.6.).

§0.1. The three problems of possibility

This book will deal with the theme of possibility, beginning with its central
philosophical problem, which will be denoted The Problem of Possibility (and,
henceforth, abbreviated as PP):

PP: What does it take for a certain situation to be possible?

For instance, what does it take for it to be possible that tomorrow it will rain
in Manhattan?
The PP is a more complex problem than it might appear to be at first. In
order to tackle it, we will make use of three sub-problems related to possibility,
which pertain to three distinct areas of research: semantics, epistemology,
and metaphysics. Semantics studies the meaning of linguistic expressions,
including the conditions under which an utterance is true or false. (For an
introduction to semantics, see Soames (2010).) Epistemology concerns the
modalities through which a subject acquires information about himself, what
surrounds him, and whatever else the subject can conceptualize—for example,
theoretical knowledge that does not concern spatio–temporal reality, or even
fantastic imaginings. (For an introduction to epistemology, see Feldman (2002).)
Metaphysics, finally, concerns entities themselves: whether they exist and
what they are, as well as how this can be explained; what type of relationships
can exist between them—for example, existential relationships, causal relation-
ships, relationships of composition between part and whole; and what modes
of classification there are. (For an introduction to metaphysics, see Loux (2006).)
The PP is therefore intimately tied to The Semantic Problem of Possibility
(SPP), The Epistemic Problem of Possibility (EPP), and The Metaphysical
Problem of Possibility (MPP):

SPP: What does it mean to say that a certain situation is possible?

EPP: How do we come to know that which is possible?
MPP: What sort of entity is a possible entity (of any given kind—a
possible individual, property, state of affairs, or …)?

9781472524263_txt_print.indd 3 01/12/2015 09:44

 4 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

The SPP, EPP, and MPP deserve special attention because possible situations
play a key role in our lives. To cite three examples: probability is nothing but a
type of possibility; therefore, when we are deciding how to invest our savings,
what course of studies to undertake, or whether to approve the construction
of a nuclear power plant in our neighborhood, we are also relying upon a
certain representation of possibility. Sentiments make for another good case
in point. Many of them, even the most visceral or uncontrollable emotions,
concern possible situations: we find ourselves fearful of having an accident,
or worried about the health of a loved one, despite the fact that these are only
possibilities, not matters of fact. Finally, the ways in which punishments and
rewards are accorded often rely upon what a person could have done or could
do: we might think, for example, of a sentence given for attempted homicide
or a promotion awarded for a person’s fulfillment of her potential. Therefore,
reflecting upon what possibility is is central to understanding ourselves, the
universe, and the norms and values that influence us, and those around us.
The objective of this book is to examine the solutions of the PP that have
been proposed, with special attention to their metaphysical underpinnings.
To this end, the SPP, EPP, and MPP will be discussed from time to time.
Before proceeding, we should consider the relationship between these three
problems.
The contemporary debate on modality developed from an important
contribution to the SPP, namely so-called possible-worlds semantics, which
reached a mature state during the end of the 1950s and early 1960s. In
possible-worlds semantics, the truth conditions of propositions—and, in
particular, of propositions involving modal terms—are evaluated with respect
to a class of possible worlds. A proposition involving a possibility operator is
true at a world when true at a possible world, while a proposition involving a
necessity operator is true at a world when true at all possible worlds. Thus,
for instance, the proposition Foffo the cat could have eaten milk and cookies
is true at our world if there is a possible world where Foffo does eat milk and
cookies; on the other hand, the proposition Foffo the cat is necessarily a cat
is true at our world when Foffo is a cat at every possible world. Equipped
with this (apparently) simple semantic machinery, which can be modified in
numerous important respects, philosophers were able to address many old
philosophical questions concerning modality in a new guise. On the other
hand, the machinery had a not-so-small issue that became increasingly
worrisome over time: what is a possible world? This question, which will
accompany us throughout the volume, sums up one of the central problems
of metaphysics of the past fifty years.
The epistemic and metaphysical doubts concerning possible worlds
reawakened a host of vexed issues surrounding modality, which are indeed
also discussed independently of the specifics of possible-worlds semantics.

9781472524263_txt_print.indd 4 01/12/2015 09:44

 INTRODUCTION: FRAMING THE DEBATE 5

The difficulty with the EPP resides in the fact that the majority of possible
entities are not actual: they are never materialized in our universe. Therefore,
they are entities that we cannot experience directly. I can know that Foffo
the cat is crossing the street because I am present at the scene; I can know
that yesterday Foffo had milk and cookies because Elena, who was present
at the scene, told me so; but how do I know that Foffo could have had milk
and cookies again today, if instead he ate fish and potatoes? Neither I nor
any other person has ever been present at that possible scene. Perhaps this
possibility is purely an invention, a projection of our minds. Or maybe the
knowledge is an induction, or an inference based on certain empirical data:
we infer that Foffo could have eaten milk and cookies today based on the
fact that he ate milk and cookies yesterday. Or perhaps it is a deduction, an
inference based on purely logical reasoning: Foffo could have eaten milk and
cookies today because nothing contradicts our thinking this. Thus, how do I
know that Foffo could have eaten milk and cookies again today?
A strategy for unraveling the knot created by the EPP is to turn to the MPP
for help. We must first clarify what type of objects possible entities are in order
to have a point of departure for explaining how we can gain understanding
of them. If possible entities are merely fictions (as the so-called fictionalist
claims), like Hamlet, then we have an argument for claiming that they are the
product of our imagination; if they are entities that are entirely similar to those
we experience (as the modal realist maintains), then we might assert that we
become familiar with them by way of induction, endorsing their existence by
relying on a philosophical analysis.
Nevertheless, some philosophers seem to follow the opposite path: in
order to resolve the MPP, they rely upon a certain solution of the EPP. This is
the case with the agnostic theory, for example. In yet a different case, the
fictionalist opts for a certain solution of the MPP based on the advantages of
a certain solution of the SPP. Deciding which strategy to follow is a question
of philosophical methodology, which the reader will find applied in a variety of
ways throughout the course of this book.

§0.2. The sphere of discussion

In what follows, the reader will be introduced to the eight principal theories
of possibility. The focus will be on the metaphysical outlook of each theory.
However, in order to appreciate the metaphysical outlook of each view and
to understand how it may be twisted or revised, the distinction between the
SPP, EPP, and MPP must be kept in mind. We will highlight which of these
three problems we are dealing with at any given point, and whether, and

9781472524263_txt_print.indd 5 01/12/2015 09:44

 6 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

how, a proposed solution can apply to the other two. Thus, while the focus
of this volume is on the metaphysical aspects of contemporary theories of
modality, along the way the reader will find remarks that apply to epistemic
and semantic aspects of these theories too.
Yet, the array of theories that will be surveyed certainly cannot be
considered complete. The theoretical field of interest to us is restricted from
two main standpoints. First, we will examine only theories belonging to the
contemporary philosophical scene, confining for the most part the historical
predecessors to Chapter 1 (more on this also in the following section).
Second, from the standpoint of the philosophical community, we will discuss
only the works of philosophers who—as they say—have an analytic approach
to philosophy (see Chase and Reynolds (2010)).
By implementing this last restriction we do not intend in any way to
embrace the view that the analytic approach is superior to others, much less
the view that there is one approach that stands out as being analytic. Rather,
we are simply advising the reader of the brute fact that the philosophical
theories considered here were produced within a circle of philosophers
belonging to a given philosophical milieu. Though the latter is not readily
identifiable, two of its traits are plain. First, by and large, the theories were
initially advanced by philosophers working in institutions in the United States,
the United Kingdom, Australia, or New Zealand. Second, the articles and
books that fuelled the debate were all originally published in English, with
the exception of a few contributions in European languages, such as German
(especially important for early stages of the debate) and Italian (one version
of the new actualism, dispositionalism, was originally published in Italian).
The geographic and linguistic aspects alone, however, are not sufficient
to pin down the content of the analytic approach to modality. As the label
itself suggests, the approach relies crucially on philosophical analysis, which
is often carried out by means of a formal language, such as quantified modal
logic, along with a possible-worlds semantics. It is part of such an analysis
to divide the philosophical problem of possibility into sub-problems and to
define the different views on the market based on the semantic/epistemic/
metaphysical treatments of modal expressions that they propose. This is
indeed the approach endorsed in this book.
Finally, it is also important to clarify that several key figures involved in the
debate were educated outside of the United States, the United Kingdom,
Australia, and New Zealand, and that they were non-native English-language
speakers. Moreover, the contemporary geographic distribution of academics
contributing to the debate on modality is much different from the distribution
of fifty years ago: it spans the five continents and dozens of countries,
including Argentina, Austria, Belgium, Brazil, Canada, China, Czech Republic,
Finland, France, Germany, Greece, Holland, Hungary, India, Israel, Italy,

9781472524263_txt_print.indd 6 01/12/2015 09:44

 INTRODUCTION: FRAMING THE DEBATE 7

Japan, Korea, Mexico, Norway, Poland, Portugal, South Africa, Spain, Sweden,
Switzerland, Turkey, and United Arab Emirates.

§0.3. Methodology
At this point, we shall mention three ways in which the PP will not be
discussed.

1 This book will not deal with the PP by making use of a formal language.

Relatively speaking, a language is natural if it can be used ordinarily and

learned without a didactic aid: examples include English, French, Chinese,
Swahili, Tagalog, and so on. By contrast, a formal language is a non-natural
language whose alphabet and grammar are entirely different from those of
English. For example, HTML-programming language is a formal language.
Logical languages are formal languages as well.
Most of the considerations in the metaphysics of possibility (in analytic
philosophy) were developed in connection with the growth of a field of
logic known as modal logic: possible-worlds semantics is a chapter—by and
large, a foundational one—in contemporary modal logic. Modal logic has
provided a conceptual system of unparalleled rigor and precision, in which
classic questions related to possibility—such as the question of free will—are
reformulated and reconsidered. In other words, the development of modal
logic has allowed the very concepts of “possible” and “necessary” to be
analyzed in a new, more accurate language.
Some of the theories examined here are rooted in modal logic, particularly
those that make reference to the concept of possible worlds. However, in this
book the reader will not find formal expressions in possible-worlds semantics,
for example, or formal expression of the modal skepticism of Willard van
Orman Quine. Both topics will be dealt with, at the proper time, in the English
language.
This decision has been made for reasons concerning space and complexity,
and due to the conviction that the central philosophical questions can (and,
at least at an introductory level, should) be discussed independently of their
representation in a formal language. When formalisms are introduced, the
discussion is often diverted from the original problem: the problem of possi-
bility. The problem is embedded in daily life: it is expressed, in our case, in
the English language; and it is our duty to give a response in the English
language.
This is not meant to discourage the reader from studying logic, and in
particular modal logic: as will be emphasized on multiple occasions, drawing

9781472524263_txt_print.indd 7 01/12/2015 09:44

 8 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

upon a formal language produces a sharpness and clarity that are unknown
in natural language. Logic remains an instrument that is to be used to clarify
a philosophical question or sharpen a solution that has been devised, albeit
in its preliminary stages. Here, however, we can rely upon decades of sharp-
ening of modal logic: the problem and its solutions have been provided in
both natural and formal languages. The main issue is which theory to choose,
and our goal is to assess the metaphysical proposals on the market.

2 This book will not deal with the PP by undertaking an historical

analysis of its solutions.

We certainly are not the first to consider the PP, and there is much to be
learned from those who have preceded us. Those who wish to gain a deeper
understanding of the concept of possibility ought to study the classic texts
by authors such as Aristotle and Leibniz. For this reason, indeed, we have
reserved an entire chapter—Chapter 1—to discuss prominent philosophical
results in modality. Yet possible-worlds semantics equips us with conceptual
resources of unprecedented sophistication, leading to a debate that deserves
to be discussed separately. More importantly, we seek to use the most
appealing theoretical apparatus for dealing with modalities, since our priority
is not historical development or accuracy but theoretical soundness. The
main ideas and questions raised in the past centuries will be kept in the
background, to guide us from time to time, especially—as we shall see in
Chapters 7 and 8—in connection with some recent and important proposals
concerning the nature of possibility and necessity.

3 This book will not deal with the PP by focusing on the classic themes
and problems associated with possibility.

The history of modality is by and large the history of theoretical apparatuses

developed to address three key problems in ethics, metaphysics, and episte-
mology/philosophy of mind, namely: to show the compossibility of free agency
in a predetermined world; to demarcate essential from accidental properties
of individuals (e.g. to be able to separate what is essential from what is
accidental to being human, or so-called natural kinds, such as electrons and
stem cells); to distinguish between propositions that pick out their referents
by description from those that—so to speak—embed their referents directly
(e.g. “There is a fat cat on the couch” vs. “Foffo is on the couch”). Over time,
the theoretical apparatuses dealing with these problems became increasingly
systematic. Possible-worlds semantics is one of the latest tools to be able to
address them, along with other cognate and remote (but no less important)
problems. Our focus will be on the theories through which we can take on all
these problems. For this reason the reader will find the problems discussed

9781472524263_txt_print.indd 8 01/12/2015 09:44

 INTRODUCTION: FRAMING THE DEBATE 9

on multiple occasions (and further readings are suggested at the end of this
Introduction), but without the systematic display they deserve.
Furthermore, it should be noted that we will be interested in what is
possible tout court, while keeping the distinction between the various
categories of possible situations in the background and pointing them out
where most suitable. Nevertheless, such categories are highly relevant in
certain areas. For example, a great deal of the literature on alethic modalities
developed in conjunction with the literature on temporal modalities. The
distinction between that which was possible, that which is possible, and that
which will be possible is at the root of how we think about agency and respon-
sibility, and serves to introduce the so-called problem of future contingents.
According to this problem (addressed in Chapter 1), just as it seems intuitive
(and in keeping with the linguistic details of most natural languages) that the
present and the past are fixed, it is equally intuitive that the future is open.
Thus, while it is a fixed truth that—say—Foffo had milk and cookies yesterday
and that Foffo is having milk and cookies now, it is still open whether Foffo
will be eating milk and cookies tomorrow—that is, both “Tomorrow Foffo will
be eating milk and cookies” and “Tomorrow Foffo will not be eating milk and
cookies” are possible futures at the moment; and yet, tomorrow only one of
the two possible futures will be actualized, and the other will be regarded
as a counterfactual situation, a situation that runs counter to the facts of the
actuality. In short, intuitions suggest that there is only one past and only one
present, but there are many futures. Needless to say, philosophers have
challenged these intuitions on many counts, and to do so they had to resort
not only to a theory of time but also to a theory of possibility. The theories we
will discuss in the book can be (and have been) used to frame and solve the
problem of future contingents in different ways; yet, it is not the aim of the
book to discuss the details of such potential solutions.
Another relevant distinction among categories of possible situations is that
between nomic possibility and metaphysical possibility. The former concerns
the laws of nature, or the behaviors that seem to be inherent to spatio–
temporal objects, while the latter concerns all of the non-contradictory ways
objects can present themselves. The distinction is central to the philosophy of
science: according to some authors (e.g. David Armstrong), what is nomically
impossible is impossible tout court; others (e.g. David Lewis) contend that
the realm of the possible is much wider than we can even imagine, extending
far beyond what does not accord with the laws of nature. The distinction
between nomic and metaphysical possibility will be considered in different
chapters (especially Chapters 4, 5, and 8). Nonetheless, it is beyond the
scope of this book to deal systematically with it.

9781472524263_txt_print.indd 9 01/12/2015 09:44

 10 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

§0.4. A bit of vocabulary

As in every field of intellectual research, philosophical reflection on possibility
has assumed certain characteristic features as well as its own terminology. In
addition to guiding the reader through the answers to the question regarding
what is for something to be possible, we will also try to familiarize her with
the distinctive terminology and concepts of the field.
To this end, the book will make extensive use of the expression “concept”,
which here refers roughly to that which can function as the subject of a
cognitive act, or mental act, through which reality is understood. For an intro-
ductory and in-depth examination of the theme, see Carey (2009).
Moreover, in natural language we use an assortment of expressions to speak
of that which is possible and necessary: “to be able to”, “to have to”, “to have
the ability”, “to have the opportunity”, “possibly”, “duly”, “perhaps”, “maybe”, and
so on. From a syntactic point of view, these are expressions that are applied to
a sentence or to a part of a sentence (for example, the subject or the predicate).
These are the adverbs, verbal modes, and enunciative modal operators
classified in technical jargon as alethic modalities (or, simply, modalities).
The term “modality” comes from the fact that a modal expression charac-
terizes the mode of existence of the entities that are referred to by the
expressions to which it is applied. Let us take, for instance, the sentence
“It is raining” along with the following variants: “Today: it is raining”; “In
Manhattan: it is raining”; “I claim that: it is raining”; “It is (morally) good that:
it is raining”; “Necessarily: it is raining”; “Possibly: it is raining”. Each of these
variants expresses a way of being of raining, that is to say, of the entity to
which “It is raining” refers.
At this point, it becomes somewhat natural to classify the various modes
into different categories, and consequently the modalities that express them.
Temporal modalities (“today”, “now”, “in 1977”, et cetera) specify the way
of being in time of the entities which are referred to by the expressions to
which they are applied. Epistemic modalities (“to believe that”, “to know that”,
“to see that”, et cetera), from the Greek επιστημη (epistéme), “knowledge”,
specify the way that the entities referred to by the expressions to which
these modalities are applied are thought of. Deontic modalities (“it is good
that”, “it is wrong”, “it should be encouraged”, et cetera), from the Greek δεομ
(déon), “obligation”, specify the “way of being moral” of the entities referred
to by the expressions to which they are applied. Finally, alethic modalities,
from the Greek αληθεια (álétheia), “truth”, specify the “way of being true” of
the entities referred to by the expressions to which they are applied.
Possibility and necessity are alethic modalities. The modality of what is
possible is used to express the fact that the existence of an entity can be

9781472524263_txt_print.indd 10 01/12/2015 09:44

 INTRODUCTION: FRAMING THE DEBATE 11

realized; the modality of what is necessary is used to express the fact that, in
any case, it will be realized.
Modality is one of the most fertile areas for philosophical reflection. It is so
fertile that it forces us to create a restriction: our discussion will be limited to
the alethic modalities. Henceforth, when we will speak of modalities, we will
be referring to alethic modalities.
These limits notwithstanding, it is unlikely for a theory of possibility that
makes contemporaneous use of all the alethic modal expressions to attain
sufficient conceptual clarity. It has, therefore, become standard to make a
choice that favors a restricted number of paradigmatic expressions of (alethic)
modal discourse, to which all of the others can in some way be connected;
such expressions are gathered within the following two groups:

1 “it is possible that” and “it is necessary that”;

2 “necessarily”, “possibly”, “accidentally”, “essentially”, “impossibly”,

“contingently”.

The modalities in group (i) are enunciative operators, that is, expressions
that modify the meaning of an entire sentence, such as “It is possible that:
tomorrow it will rain in Manhattan”. The modalities in group (ii), on the other
hand, are adverbs, expressions that modify the way of existence of certain
entities, be they individuals, events, properties, relationships, or concepts;
an example is the use of “necessarily” in: “Cheese is necessarily produced
from milk”. In one way or another, all of the theories presented here
presuppose that the expressions in groups (i) and (ii) are paradigmatic:
based on these we are able to shed light on many other (alethic) modal
expressions in our ordinary language; the theoretical challenge, then, lies
in revealing the meaning of the expressions gathered in (one of) these
groups.
It would be equally useful to alert the reader to the following terminological
choices that have been used in the text.

● “Theory of possibility”/“Theory of modality”: These expressions

are used synonymously: a theory of possibility, just like a theory of
modality, must explain not only the meaning of expressions such as
“it is possible that” or “possibly”, but also expressions such as “it is
necessary that” or “necessarily”.
● “Ordinary language”/“Natural language”: These expressions are
also used as synonyms, to designate any language that can be
learned without a didactic aid and used ordinarily. Languages used
to formulate any formal logic are not to be understood as natural or
ordinary.

9781472524263_txt_print.indd 11 01/12/2015 09:44

 12 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

● “Situation”: This expression stands for what is designated or

described by a complete sentence, no matter its form: an action,
an event, a state, or a relationship among these. In contemporary
analytic philosophy, the expression “state of affairs” is often used
rather than “situation”. But because the former is now associated
with specific metaphysical positions about which we wish to remain
neutral, we have chosen to use the latter.

§0.5. Classification of theories and order

of exposition
Most philosophers are inclined to consider their position to be distinct from
those of others. And given that the philosophers who have dealt with possi-
bility number well over eight (the number of theories we will discuss), it is
very difficult to deliver an agreeable description of the debate in the chapters
that follow. Texts that provide a comprehensive overview of the debate
about the metaphysics of modality do vary to some extent with respect to
the number of theories they discuss, and the names of the theories. This
book is organized around eight main theories, that is, eight philosophical
responses to the PP and its three sub-problems SPP, EPP, and MPP. The
classification adopted strikes us as effective and easy to remember, and as
carving the debate at three crucial joints. In the remainder of this section we
will introduce the classification of the theories by illustrating the three philo-
sophical problems that determine said crucial joints. For an illustration of the
classification, see Figure 0.1. (As anticipated, Chapter 1, which is not in the
diagram, deals with the history of work on modality.)

§0.5.1. A genuinely conceptual dispute?

The first joint dividing philosophers on matters of modality lies at the inter-
section between the cognitive and the non-cognitive. Do modal expressions
really express concepts? Are we able to grasp these alleged concepts? The
wide variety of modal expressions we find in most natural languages, the
abundance of purposes they serve, and, most importantly, the opacity of
their roles invite us to answer these questions in the negative. Indeed, some
deny that alethic modalities express concepts: either because we are not
able to truly understand what we speak of (modal skepticism), or because
what is at stake are sentiments, not concepts (modal expressivism). Modal
skepticism and modal expressivism will be taken up in Chapter 2.

9781472524263_txt_print.indd 12 01/12/2015 09:44

 INTRODUCTION: FRAMING THE DEBATE 13

The vast majority of philosophers, however, maintain that the task of a

theory of possibility is precisely to explain certain concepts. Against the
skeptics we can point out that modal expressions are constantly employed
where conceptual rigor seems to be most desirable and logics dealing with
them abound. Against the expressivists we can point out that modal expres-
sions can be nested, to express subtleties that typically only concepts can
exhibit; David Lewis offers a nice example (borrowed from a 1970 paper by
J. Howard Sobel) in his book Counterfactuals (1973):

If the USA threw its weapons into the sea tomorrow there would be war;
but if the USA and the other nuclear powers all threw their weapons into
the sea tomorrow there would be peace; but if they did so without suffi-
cient precautions against polluting the world’s fisheries there would be
war; but if after doing so, they immediately offered generous reparations
for the pollution there would be peace. (10)

Whether those and other rejoinders are sound, the expressivist position
seems to defend an important intuition, which deserves to find room in a map
of theories of possibility.

§0.5.2. A conceptual reduction?

The second joint in the debate on the metaphysics of modality divides those
who maintain that alethic modalities cannot be analyzed in the terms of
other concepts (the modalists) from those who maintain that they can. This
is a recurring kind of distinction in analytic philosophy: analysis must have an
end, but it is disputable where to draw the end line. Modalists—whose view
will be discussed in Chapter 3—hold that there is no clearer way to explain
what it means to say that Napoleon could have won at Waterloo than to
say … Napoleon could have won at Waterloo. To clarify, prominent modalists
such as Graeme Forbes and Charles Chihara do believe that the variety of
modal expressions can be captured in terms of two chief ones—that is,
necessity and possibility; they also believe that the meaning of expressions
aiming to convey the necessity or possibility of a situation is best captured
by a formal language. Yet they deny that possible-worlds semantics permits
us to infer the existence of possible worlds, regardless of what these may
be taken to be. The meaning of necessity and possibility operators is not
analyzable.
Others, however, have attempted to explain modal concepts. To the
credit of modalism, the trend of analyzing modal concepts in terms of—
say—possible worlds or dispositions is rather recent. Talk of possible worlds

9781472524263_txt_print.indd 13 01/12/2015 09:44

 14 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

can be found far back in the history of philosophy, such as in the writings
of Epicurus (341–270 BCE). But it was not until Leibniz (1646–1716) that the
concept was used to systematically analyze propositions containing modal
expressions. And again, until the 1960s, when possible-worlds semantics
was more firmly established, modalism seemed the most prudent view.
However, matters are now different: today most philosophers working on
modality hold that modal expressions should be analyzed in terms of other
concepts, possible worlds being chief among these. The widespread debate
about the semantics, epistemology, and metaphysics of modality generated
over the past fifty years may, indeed, be considered an attempt to fine-tune
such alleged conceptual analysis.

§0.5.3. Possible worlds?

Among those who have attempted to explain modal concepts, some (modal
realists, ersatzists, fictionalists, agnostics) make use of possible-worlds
semantics. Modal realism, which will be taken up in Chapter 4, became
prominent thanks to the writings of David Lewis, in particular the book On
the Plurality of Worlds (1986), now a classic in contemporary metaphysics.
According to Lewis, there is an infinity of worlds, each of which is as real or
concrete as our world is. No two worlds overlap, though (parts of) worlds may
resemble each other in virtue of their inhabitants sharing properties or taking
part in the same events.
Lewis’s view—which, in his own words, has often prompted an “incred-
ulous stare”—was criticized left and right. The largest contingent of critics
belongs under the label of “ersatzism”, the subject of Chapter 5. The view
owes its name to the German “ersatz”, roughly translated as “surrogate”. For
ersatzists, possible worlds are surrogates of the actual world; unlike modal
realists, they regard the actual world as having metaphysical priority over the
other worlds. There are many versions of ersatzism, some of which can be
traced back to Rudolf Carnap, Ludwig Wittgenstein, or even Leibniz. In On the
Plurality of Worlds, Lewis raises some serious objections against this view,
which he considers as the major rival on the market.
Shortly after the publication of Lewis’s landmark book, Gideon Rosen, a
colleague of his at Princeton, put forward a novel view, modal fictionalism,
which will be discussed in Chapter 6 alongside the more recent modal agnos-
ticism. There are a few variants of fictionalism, some of which have gained
credence in recent years. All variants agree in the fundamental strategy, which
can be divided into two stages: first, take the most conceptually powerful
view on the market, modal realism, and pretend that it is true only for the
sake of conceptual analysis; second, discard any ontological commitment of

9781472524263_txt_print.indd 14 01/12/2015 09:44

 INTRODUCTION: FRAMING THE DEBATE 15

modal realism—it is merely a fiction. As we shall see, this strategy is difficult

to pin down and probably lacks some of modal realism’s theoretical virtues.
Modal agnosticism, on the other hand, was first proposed by John Divers
in a 2004 article, “Agnosticism About Other Worlds”. As the name indicates,
Divers suggests that we take possible-worlds semantics at face value, but
remain agnostic as to the existence of possible worlds (regardless of what
these may be taken to be). The details of this position are still evolving and
under discussion; we will consider some objections in Chapter 6.
Finally, we come to those who reject possible worlds, but still aim to
provide a conceptual reduction of modal expressions. These philosophers
largely overlap with those whom Vetter (2011a) labels “new actualists”, so
we will borrow the term. New actualism will be analyzed in Chapter 7; it is
the most recent position on the map and it is still under scrutiny and devel-
opment. We will distinguish two main variants: the first (dispositionalism)
relies on dispositions, entities’ abilities to act a certain way if placed in certain
conditions; the second (new essentialism) insists on the importance of
preserving multiple, irreducible modal expressions—essence, metaphysical
necessity, nomic necessity, and deontic necessity. Both positions have
generated a considerable amount of interest in recent years, and lie at the
core of prominent ongoing debates in modality. Dispositionalism was
first suggested by Ellis (2001) and Molnar (2003), more explicitly defended
first by Borghini (2003) and then by Borghini and Williams (2008), and later
refined and developed in a number of works, including those of Contessa
(2010, 2013), Mumford and Anjum (2011), and Vetter (2015). Although thinkers
such as Aristotle recognized the importance of dispositions, the attempt
to analyze the wide range of modal expressions in terms of dispositions
is a bold one; and, as we shall see, it has been subject to some important
criticisms. The new essentialism owes much to the works of Kit Fine (e.g.
2005). In some respects, Fine’s position is akin to modalism (see for instance
Melia (2003: Chapter 4)); this is because according to Fine there is a core of
modal concepts that is ultimately irreducible, just as the modalist claims that
necessity and possibility operators express irreducible concepts. However,
Fine’s work has also generated novel literature on such topics as essences
and grounds, thus prompting new conceptual analyses of modal expressions.
For this reason, new essentialism is classified alongside new actualism and
not modalism.
While possibility and necessity are logically and hence theoretically
entrenched, there are some questions pertaining to the truth of necessary
statements that deserve to be discussed independently of specific
theories of possibility. This is the task of Chapter 8, which will discuss the
varieties of necessity (pragmatic, nomic, deontic, conventional, metaphysical,
and essential); necessary truths concerning mathematical entities, such

9781472524263_txt_print.indd 15 01/12/2015 09:44

 16 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

as geometrical figures and numbers, as well as truths concerning abstract

entities such as sets; and, finally, a priori truths and the epistemic accessi-
bility of modal truths.

§0.6. Impossibility
It is our intention to look carefully at certain theories of possibility. Yet—we
might ask ourselves—what can be said about impossibility? Once we have a
theory of possibility, will we also be able to account for that which is impos-
sible? Yes and no. Once we establish an interpretation of alethic modalities,
we will be able to tell where the boundary lies between that which is possible
and that which is impossible. However, this does not mean that we will know
how to treat the questions that emerge once impossible situations are taken
into consideration: if two plus two were to equal three, then could I also go
surfing this afternoon on twenty-seven different surfboards simultaneously?
Or, could we admit a given impossibility without compromising the truth of
the other things we think?
Since the contributions of Morgan (1973) and Yagisawa (1988), respectively
addressing logical and metaphysical issues regarding impossible worlds,
these and cognate questions have become increasingly prominent in the
literature on modality. Impossible worlds have been particularly handy in
explaining the behavior of agents who hold inconsistent beliefs, as well
as in addressing the truth conditions of systems of propositions (e.g. legal
systems) that contain contradictory claims. We shall come back to this topic
in Chapter 8.

Study Questions
• What is a modality a mode of?
• What are alethic modalities?
• What other modalities, besides the alethic, are there?
• What is the problem of possibility?
• What are the three main sub-problems of possibility?
• What is possible-worlds semantics?
• What are the three main joints giving shape to the contemporary debate
on alethic modality?
• What is an impossible world?

9781472524263_txt_print.indd 16 01/12/2015 09:44

 INTRODUCTION: FRAMING THE DEBATE 17

MODAL EXPRESSIONS

EXPRESS CONCEPTS?
SKEPTICISM
NO YES

EXPRESSIVISM
CONCEPTUAL REDUCTION?

MODALISM NO YES

POSSIBLE WORLDS?

NO YES

DISPOSITIONALISM AND NEW

ESSENTIALISM

MODAL REALISM

AGNOSTICISM

ERSATZISM
FICTIONALISM

FIGURE 0.1 Classification of the theories of possibility adopted in the volume

9781472524263_txt_print.indd 17 01/12/2015 09:44

 18 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

FURTHER READING
For an introduction to modal logic, see Hughes and Cresswell (1968, 1984),
Chellas (1980), Fitting and Mendelsohn (1998), Girle (2000), and Garson (2006).
The last is particularly suggested for the more philosophically and less logically
oriented.
On the history of modalities, see the classic by Kneale and Kneale (1962).
For a reconstruction of the main moments in the development of possible-worlds
semantics, see Copeland (2002) and Ballarin (2010); among the essential readings,
see C. I. Lewis (1918), Wittgenstein (1921), C. I. Lewis and Langford (1932), Carnap
(1946, 1947), Prior (1957), Kripke (1959, 1963), and Hintikka (1961, 1963).
For an overview of the debate regarding modality and free will, see Kane
(2002). Among the most influential classic contemporary texts on this topic are
Plantinga (1974b), Pike (1977), van Inwagen (1983), and Vihvelin (1988).
The debate on essentialism traverses the whole history of philosophy and is
among the most assorted and vast; for a first approach, see Robertson (2013).
The readers, edited respectively by Linsky (1974) and Loux (1979), are another
useful starting point. The most significant contemporary texts on the topic include:
Kripke (1980), Wiggins (1980), Sidelle (1989, 2002), Ellis (2001, 2002), Fine (2005),
Mackie (2006), and Lowe (2006). For an introductory reading on natural kinds,
see Bird (2009) and Bird and Tobin (2015). Beebee and Sabbarton-Leary (2010)
and Campbell, O’Rourke, and Slater (2011) are updated collections of essays on
metaphysical and semantic issues concerning natural kinds; among the recent
monographic studies, see LaPorte (2009) and Khalidi (2013).
For a historical debate on future contingents, see the classic by Gaskin (1995),
and the earlier classic work by Prior (1957). Hasle and Øhrstrøm (2011) provide an
up-to-date framing of the debate.
For additional introductory texts on contemporary theories of possibility in
analytic philosophy, see: Divers (2002), Melia (2003), Loux (2006: Chapter 5), and
Pruss (2011). Also useful, although covering the debate only in part, is Sider (2003).
There are several readers collecting important writings about modality. Among
these are: Linsky (1974); Loux (1979); Gendler and O’Leary Hawthorne (1992); and
Hale and Hoffman (2010).
Finally, for an overview of the main issues relating to impossible worlds, see:
Yagisawa (2010); Berto (2013a); and Jago (2014). See also the Further Reading
section in Chapter 8 for additional bibliographical resources.

9781472524263_txt_print.indd 18 01/12/2015 09:44

 1
The metaphysics of modality:
A historical overview

T his chapter examines the key ideas and principal figures that, in the history
of Western philosophy, have characterized and shaped the central theories
of modality. It is divided into three main parts: Ancient (§1.1.), Medieval (§1.2.),
and Early Modern (§1.3.). After considering the early ideas of Parmenides and
Zeno (§1.1.1.), we examine the first author to have systematically studied modal
concepts, Aristotle (§1.1.2.), and the first school to have done the same, the
so-called Megarian School (§1.1.3.). The period in which reflection on modality
flourished, however, is the Medieval, with the Arabic and Scholastic philosophical
traditions (§1.2.1.). During the late Scholastic period, authors such as Ockham
(§1.2.2.) and Buridan (§1.2.3.) enriched and refined the Aristotelian and Megarian
teachings, setting the stage for the contemporary revival of reflection on modality.
During the Early Modern period, modalities were used to put forward some
bold metaphysical views, such as those of Descartes (§1.3.1.), Leibniz (§1.3.2.),
Spinoza (§1.3.3.), and Hume (§1.3.4.). The concluding section (§1.4.) highlights the
importance of nineteenth-century logic for the rise of quantified modal logic and
possible-worlds semantics in the twentieth century (examined further in Chapters
2 and 3).

*****

T he ability to reason using modal notions is characteristic of humans,

and is certainly remarkable, arguably playing an important role in our
evolution and setting us apart from other intelligent beings. Musil’s famous
passage on the sense of possibility, cited in the Introduction (see page 1),
shows concisely the centrality of modal intuitions to human endeavors.

9781472524263_txt_print.indd 19 01/12/2015 09:44

 20 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

These intuitions often play an important role in assessing the palatability of

a theoretical option and in getting it off the ground, and during our journey
through the theories of possibility we shall have several occasions to appre-
ciate this role. It is no surprise, then, that modality has been discussed since
the earliest historical examples of philosophical speculation (see Gabbay and
Woods (2004a)).
We shall begin our study of the metaphysics of modality with an excursus
through the history of philosophy, examining the main authors and arguments
that influenced Western theories of possibility and necessity. While the story
we will tell starts with Ancient Western philosophy, it is important to note
that its roots run much deeper into human history: in the work of key authors
such as Parmenides and Aristotle, we can trace important influences from
Ancient Indian philosophy and Jewish thought via the Babylonian, Egyptian,
and Cretan civilizations (Gabbay and Woods 2004a). We shall not trace these
connections, but rather stick to those texts and authors that more directly
motivated the texts and also writers who take center stage in the chapters to
come.
The discussion to follow has been organized around three main moments:
Ancient (§1.1.), Medieval (§1.2.), and Early Modern (§1.3.). It goes without
saying that there is an element of artificiality in this schema, and that the
development of ideas about possibility and necessity follows not time, but
variables such as place, school, cultural context, and individual genius. Our
tour through this history will include only a few stops, and only for work that
is so classic as to be essential to even an introductory treatment; and we shall
give voice to the original texts as much as possible. The views of each author
we examine have been the object of meticulous and extensive contemporary
study, and often of complex controversies of interpretation; the reader can
find ample suggestions for introductory presentations, monographic studies,
and research articles in the Further Reading section at the end of this chapter.
For each author, we shall provide—where possible—dates of birth and death
as well as key locations where the author is known to have lived and been
active.

§1.1. Modalities in Ancient philosophy

§1.1.1. Parmenides and Zeno
Parmenides (c. 515–450 BCE; Elea, Campania, Italy) is often considered to be
the first significant figure in Western thought. Only about 160 fragmented
lines of the original 3,000 lines of his philosophical poem “On Nature” are

9781472524263_txt_print.indd 20 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 21

known to us, and yet based on these fragments, and also on the discussion of
Parmenides’s ideas in the work of other authors (such as Zeno and Diogenes
Laertius), scholars have been able to agree about some of the central tenets
of his philosophy.
In the second part of “On Nature”, Parmenides lays out his chief metaphysical
thesis: Being is all that there is, while not-being cannot be. Superficially, this
may appear an uncontroversial claim; far from it, however, since Parmenides’s
thesis is a formulation of so-called metaphysical monism: there is only one
thing, Being, which never changes and so never dies, which never came into
existence, which has no boundaries and no detachable parts. Metaphysical
monism contends not only that the entire universe is one simple entity,
but—most importantly for our purposes—that this entity has some necessary
features: it cannot go out of existence (it is eternal) and it cannot change.
Parmenides may have defended, or have come close to defending, what we
now call necessitarianism. This is composed of two theses: (i) the universe
exists out of necessity, and (ii) it is necessarily the way it is. Metaphysical
monism is an endorsement of (ii); however, whether Parmenides also
endorsed (i) remains controversial (see Lewis (2009)). We shall encounter
an instance of necessitarianism later on in this chapter when discussing the
views of Baruch Spinoza (1632–77).
Possibility, as well, plays a role in Parmenides’s philosophy. While true
knowledge is recognizing the necessity of Being, false opinion is that
which suggests the contingency of worldly affairs—such as that things
could have been otherwise or that the future is open. It seems plausible
that Parmenides’s philosophy must be read as a reaction to Pythagorean
science, which tried to explain worldly phenomena in terms of a multiplicity
of opposing principles. Parmenides’s method emphasized the importance of
reason over the senses: philosophy becomes that way of life which goes
beyond the surface of appearances and reveals, by means of the use of
reason, the truth of seemingly incredible conclusions.
The Parmenidean reliance on reason is also at play in the work of his
student (or young associate) Zeno of Elea (Campania, Italy), active during the
fifth century BCE. Zeno is most famous for his ingenious paradoxes, having
indeed introduced this genre of work in Western philosophy. It seems that the
goal of Zeno’s arguments was to lend credence to Parmenides’s metaphysical
monism. Confronting us with claims that at least prima facie contradict
each other, the paradoxes force us to rethink the bounds of possibility and
necessity. Zeno’s paradoxes stand as excellent examples of the importance
of modal reasoning to philosophizing. Relevant for our purposes is that they
invite us to entertain some (seemingly) impossible scenarios; we shall come
back to the importance of impossible scenarios for the metaphysics of
modality at the end of Chapter 8.

9781472524263_txt_print.indd 21 01/12/2015 09:44

 22 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Eight paradoxes have survived, although not all in their original formulation,
They are typically divided into four groups: four paradoxes against motion;
two paradoxes against plurality; the paradox of the millet seed; and the
paradox of place. Let us briefly go through each of them.

§1.1.1.1. The four paradoxes of motion

While their original formulation is lost, we find these paradoxes discussed
in the works of other philosophers, including Aristotle’s Physics. The lack of
textual evidence adds to the interpretive difficulty, as different readings of the
paradoxes are hinted at by different formulations. Zeno’s aim seems to have
been to show the absurdity of Pythagorean science, which was atomistic—it
explained worldly phenomena in terms of the interaction of atoms. The formu-
lation of the paradoxes begins by raising a question: if atoms exist in space
and time, are space and time either continuous (and hence infinitely divisible)
or discrete? To this question we can provide four answers: that both space
and time are continuous; that both are discrete; that space is continuous and
time is discrete; that time is continuous and space is discrete. Each of the
four answers is shown to be paradoxical.

1 When both space and time are discrete, we have the so-called
Paradox of the Stadium.

2 When space is discrete and time continuous, we have the paradox of

Achilles and the Tortoise.

3 When space is continuous and time discrete we have the Bisection

Paradox.

4 When both space and time are continuous we have the Paradox of
the Arrow.

The paradox of the stadium

Suppose that space and time are discrete. Imagine you have three logs of
wood—A, B, and C—lying on the floor. Each log is three feet long. The logs
are parallel to each other and are arranged in this manner: A is in the center
of the room and is standing still; B is at the far left of the room and is moving
right along its length; C is at the far right of the room and is moving left along
its length. B and C, the two moving logs, travel at the same speed and start
overlapping A at exactly the same time. There will thus be a moment when A,
B, and C are all traversing each other, but they will do so at different speeds.
In fact, in the time it takes for the front edge of B to pass one foot of A, the
front edge of C will pass two feet of B (because both B and C are moving, and
in opposite directions). Now, suppose we change the story in two apparently

9781472524263_txt_print.indd 22 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 23

uncontroversial respects, by saying that each of A, B, and C occupy three

units of the smallest unit of space, and that both B and C are moving at such
a speed that they traverse the smallest unit of space in the smallest unit of
time. Then, for B, it takes half the smallest unit of time to traverse a section of
C that occupies the smallest unit of space; and, vice versa, for C, it takes half
the smallest unit of time to traverse a section of B that occupies the smallest
unit of space. But this is absurd, because the smallest unit is by definition
indivisible. (The Paradox of the Stadium presents some particular interpretive
dilemmas; the formulation provided here follows especially Siderits and
O’Brien (1976).)

The paradox of Achilles and the Tortoise

Assume that space is discrete and time is continuous. Achilles is running
after the tortoise and there are infinite units of time between the current
moment and the point at which Achilles catches up (because we are
supposing that time is continuous). In each unit of time, the tortoise must
traverse at least one minimum unit of space, which means that the tortoise
will traverse infinite minimum units of space, since there are infinite units
of time. In order to reach the tortoise, Achilles must traverse each of the
minimum units of space that the tortoise has traversed. But those units
are infinite. Therefore, Achilles must travel an infinite distance to reach the
tortoise, which is absurd.

The bisection paradox

Assume that space is continuous and time is discrete. Imagine a person
walking from her bed to the restroom. Since time is discrete, she will take
at least one minimum unit of time to cover each minimum unit of space. But
there are infinite such units of space, so she will take an infinite time to reach
the restroom, which is absurd.

The paradox of the arrow

Suppose that both time and space are continuous. Imagine an archer
shooting an arrow. In each minimum unit of time, which is infinitely small, the
arrow will cover a minimum unit of space, which is infinitely small: thus, in
each minimum unit of time the arrow will be at rest. So, the arrow is at rest
in every moment and is never moving, which is absurd.

§1.1.1.2. The two paradoxes of plurality

These paradoxes aim to prove that there cannot be a plurality of entities.
The first paradox shows the absurdity of thinking that there are finitely many
entities. Assume that the number of entities is finite; then, you must conclude

9781472524263_txt_print.indd 23 01/12/2015 09:44

 24 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

that there are infinite entities, because between any two entities there have
to be infinitely many dividing them. The second paradox of plurality deals with
the absurdity of infinitely small and infinitely large entities. Suppose that the
universe is inhabited by infinitely small entities (e.g. the points of a line); then,
take one such infinitely small entity and add it to any plurality of entities: at
the end of the process we will have the same number as we had at the start,
because the added entity is infinitely small; that is, we added nothing. If, on
the other hand, the universe is inhabited by entities that occupy space, then
such entities will have infinite parts; hence, they will be unlimited—which is
absurd.

§1.1.1.3. The millet paradox

This paradox focuses on the joint action of multiple entities. Suppose that
there are multiple entities, say a good bunch of millet seeds; and suppose
further that those seeds perform some joint action—say, they make a sound
when dropped into an empty bucket. Then, each seed or each part of a seed
(regardless how small) must make a sound too—which is absurd, since one
tiny seed makes no sound.

§1.1.1.4. The paradox of place

This paradox focuses on the relationship between an entity—say, a chair—
which occupies a place, and the place that is occupied by the chair, where the
chair and the place are thought of as two distinct entities. If the chair has a
place, then place itself—which is a separate entity from the chair—must have
a place; but the place of a place must have a place too; and so on ad infinitum.
Thus, nothing has a place, which is absurd.
Zeno’s paradoxes stand as one of the clearest examples of the power of
rational argumentation: without relying on any specific observation, Zeno
rejects metaphysical pluralism. It is crucial for us to stress that the argumen-
tative force of these paradoxes rests on our ability to imagine scenarios and
to suppose that they genuinely represent our universe.
In the Parmenides, Plato states that Parmenides and Zeno paid a joint visit
to Socrates in Athens around 450 BCE, and it seems plausible that Zeno had
quarters in Athens for a time. The two philosophers from Elea may therefore
have influenced Socrates and, through him, Plato and Aristotle. Certainly, the
latter gave considerable attention to Zeno’s paradoxes in the Physics and,
for what concerns Plato, he touches upon problems related to modalities
in a few places (e.g. the Timaeus, where we find a thorough discussion of
necessity). But, on the basis of the textual evidence that we have, Aristotle’s
study of modalities reached far beyond the problems presented by Zeno’s

9781472524263_txt_print.indd 24 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 25

paradoxes and by other eminent colleagues, including Plato: Aristotle put

forward the first articulated treatment of possibility and necessity in Western
philosophy.

§1.1.2. Aristotle
Scholars of Aristotle have discussed at great length on his position with
respect to modalities; a selection of significant publications can be found
in the Further Reading section at the end of the chapter. In this section, we
shall restrict our focus to two tasks: first, we will distinguish the different
modalities that seem to emerge from Aristotle’s writings, asserting that
some important ones—logical possibility, synchronic possibilities, and condi-
tionals—are missing; then, we will present and discuss in more depth the
so-called problem of future contingents. But, first, let us note that another
aspect of Aristotle’s metaphysics often plays an important part in the contem-
porary debates on the metaphysics of modality: the theory of substances, and
the related doctrine of essentialism. Since essentialism will be discussed at
various stages in the chapters to come (especially Chapters 3, 7, and 8), we
shall set this topic aside for now.

Necessities and possibilities

Aristotle (c. 384–322 BCE; Stagira and Athens, Greece) examines issues with
possibility and necessity in many parts of his works, and most famously in the
De Interpretatione, the Prior Analytics, and the Metaphysics. He considers
four main topics: necessity; statistical or temporal frequency; diachronic
possibility; powers or capacities. Let us review each of them in order.

Necessity
Aristotle distinguishes several senses of necessity in a particularly telling
passage of the Metaphysics, book Delta, amidst the treatment of other funda-
mental notions such as “cause,” “nature,” and “element.” Aristotle writes:

We call ‘necessary’ (1) (a) that without which, as a condition, a thing cannot
live; e.g. breathing and food are necessary for an animal; for it is incapable
of existing without these; (b) the conditions without which good cannot
be or come to be, or without which we cannot get rid or be freed of evil;
e.g. drinking the medicine is necessary in order that we may be cured of
disease, and a man’s sailing to Aegina is necessary in order that he may get
his money. (2) The compulsory and compulsion, i.e. that which impedes
and tends to hinder, contrary to impulse and purpose. For the compulsory

9781472524263_txt_print.indd 25 01/12/2015 09:44

 26 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

is called necessary (whence the necessary is painful, as Evenus says:

‘For every necessary thing is ever irksome’), and compulsion is a form of
necessity, as Sophocles says: ‘But force necessitates me to this act’. And
necessity is held to be something that cannot be persuaded—and rightly,
for it is contrary to the movement which accords with purpose and with
reasoning. (3) We say that that which cannot be otherwise is necessarily
as it is. And from this sense of ‘necessary’ all the others are somehow
derived; for a thing is said to do or suffer what is necessary in the sense
of compulsory, only when it cannot act according to its impulse because of
the compelling forces—which implies that necessity is that because of
which a thing cannot be otherwise; and similarly as regards the conditions
of life and of good; for when in the one case good, in the other life and
being, are not possible without certain conditions, these are necessary,
and this kind of cause is a sort of necessity. Again, demonstration is a
necessary thing because the conclusion cannot be otherwise, if there
has been demonstration in the unqualified sense; and the causes of this
necessity are the first premises, i.e. the fact that the propositions from
which the syllogism proceeds cannot be otherwise.
—Trans. W. D. Ross; available at The Internet Classics Archive: http://
classics.mit.edu/Aristotle/metaphysics.5.v.html

The passage identifies a general definition of necessity, then divides it into

two kinds—conditional and absolute; conditional necessity is in turn divided
into three kinds. The general definition, provided under point (3), is as follows:
(N) Necessity is that which cannot be otherwise. Aristotle’s idea, here, seems
to come close to the contemporary logical notion of necessity, where a
sentence S is regarded as necessary when it is not possible that its negation
is true. This general definition supplies an umbrella under which to subsume
two more specific notions of necessity. The distinction is drawn at the end
of the passage, where Aristotle talks about the necessity of a geometrical
demonstration. Geometry suggests a case of absolute or intrinsic necessity.
In the case of geometry, the premises are necessary and their necessity
is transmitted to the conclusion; if we agree that the nature of a triangle
follows necessarily from its definition, then the sum of the angles of a triangle
is necessarily 180 degrees in virtue of the necessary nature of the triangle. In
deductive demonstrations (e.g. demonstrations in mathematics, geometry,
and logic), the entities featured in the premises, in virtue of their necessary
natures, secure the necessity of certain conclusions regarding them. On the
other hand, many cases of necessity seem to involve a conditional necessity,
such as the fact that eating is necessary for survival: if I have survived until
now I must necessarily have eaten, but it is not necessary that I eat, and it is
thus not necessary that I survive. Conditional necessity is divided into three

9781472524263_txt_print.indd 26 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 27

kinds, identified at the beginning of the passage. First, there is (N1), the
necessity of a material condition for the obtainment of some other condition;
for instance, eating is necessary for survival. Second, there is (N2), the
necessity of certain courses of action for achieving certain goals; for instance,
I must take a medicine in order to heal. The difference with respect to (N1)
is that, in (N2) the condition requires a will: someone must choose to act in
a certain way and that choice is necessary for obtaining the relevant goal.
Finally, there is (N3), the necessity of an action that is constricted or forced:
for instance, where we have no choice but to surrender to our opponent.
In several passages, Aristotle reflects on the relationships between what
is necessarily the case, what is possibly the case, and what is actually the
case. Setting aside some controversial points of interpretation, Aristotle
seems to accept and establish certain principles that eventually become core
to the development of modal logic and modal metaphysics: (i) if something is
necessarily the case, then it is actually the case; (ii) if something is actually
the case, then it is possibly the case; in Medieval times, this principle will be
summed up in the motto ab esse ad posse valet consequentia (it is allowed
to infer possible existence from actual existence); (iii) by transitivity then, (i)
and (ii) imply that what is necessarily the case is also possibly the case.
Despite this apparently settled net of relations between modal notions, we
do not find in Aristotle a systematic and unitary treatment of modalities. The
fragmented picture is most apparent when we examine Aristotle’s treatment
of possibility. Here we find at least three different notions: statistical or
temporal frequency, diachronic possibility, and power or capacity.

Statistical or temporal frequency

Sometimes, possibility stands for that which at some point in time will be
the case, and necessity for that which is the case at all times. For instance,
because salt dissolves in water at all times in which we merge salt (and
only salt) in water (and only water), we may conclude that necessarily salt
dissolves in water. The relationship between salt and water, however, seems
not to be a matter of conceptual necessity: arguably, there is nothing in the
idea of salt that necessitates that it dissolves in water; and some may even
think that we can easily imagine a scenario were salt does not dissolve in
water. The notion of a statistical modality traverses the whole history of
Western philosophy, and has been crucial at several points in time, including:
Late Medieval philosophy (see Knuuttila (2013)); the dispute between Kant
and Hume on the necessity of natural laws (see De Pierris and Friedman
(2013)); and the contemporary debate on the relationship between natural,
semantic, and metaphysical necessity (see Bird and Tobin (2015), Beebee and
Sabbarton-Leary (2010), and Sidelle (2002)).

9781472524263_txt_print.indd 27 01/12/2015 09:44

 28 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Diachronic possibility
The notion of diachronic possibility captures those scenarios that seem to
change modal status as time goes by. Yesterday, Maria gave Luca some kulfi
ice cream; Luca could have put the kulfi ice cream in the freezer, so that today
it would have been still good to eat; alas, Luca forgot the ice cream on the
kitchen counter, and the ice cream is irreparably spoiled now. So, yesterday it
was the case that the ice cream was edible, and today that is no longer the
case; more importantly, yesterday it could have been the case that Luca got
to eat ice cream, but today it is no longer possible for Luca to eat ice cream.
There is a notion of necessity related to diachronicity too. If something
must be the case at some point in the future, then we may say that it is
diachronically necessary. For instance, to say that I am mortal is arguably to
say that, necessarily, at some point in the future I will die. The diachronic
notion is different from the statistical notion: to say that necessarily I will die
is different from saying that, at every point in time, I will die (the latter is no
longer the case after my death); and it is different from saying that, at every
point in time, I must have died (it is obviously false now); and it is arguably
different also from saying that, at every point in time, I must have died or will
die (as I may be dying). So, the notion of possibility that is conjoined with a
diachronic perspective reaches quite deep into the semantics of our modal
talk; it is this notion that is at the root of a very difficult philosophical problem,
the problem of future contingents. We shall turn to this problem in a moment,
after examining the third notion of necessity—power or capacity.

Powers or capacities
The last notion of possibility we find in Aristotle is the one of power, sometimes
also referred to as “capacity” or “disposition”. Vipers’ poison has the dispo-
sition or capacity to kill humans, in the absence of an antidote; humans have
the power to think that things could have been otherwise; this key has the
capacity to open that door. Powers do not establish a link at a specific point
in time: under suitable conditions, it is correct to ascribe a certain power to
a certain entity, even if that entity has never displayed the power. Maria has
the capacity of saying ten times in a row the expression “Ten times in a row”,
even if she has never displayed it and—as is most likely—she never will. Note
that there is a link between diachronic possibility and powers: some powers
can be gained and lost with time, for they are tied to certain other abilities.
Thus, while certain entities, such as ice cream, spoil forever, other entities
can lose a certain capacity and then gain it back. For instance, when she used
to train, Maria could run for 15 miles straight; now she cannot, but she plans
on regaining that ability with two months of training. Aristotle’s discussion of
powers is quite important to the contemporary debate on the metaphysics

9781472524263_txt_print.indd 28 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 29

of modality (see for instance Marmodoro (2015)). In Chapter 7, we shall return

to powers, capacities, and dispositions, when discussing the so-called dispo-
sitional theory of possibility.
Aristotle’s writings offer a wealth of ideas from which later theorists of the
metaphysics of modality have drawn inspiration. At every point in the history
of Western philosophy, Aristotle’s ideas have been invoked and developed in
different directions. Yet, for our purposes, it is worth noting that some notions
of possibility that are important to the contemporary debate are absent from
Aristotle’s works. Three such notions shall be pointed out: logical possibility,
synchronic possibility, and conditionals.

Logical possibility
Aristotle does not really entertain the idea that to be possible is to be logically
non-contradictory. For instance, there seems to be nothing contradictory—
one could argue—in Maria’s being one-third human, one-third Alaskan
salmon, and one-third oak tree. And yet, such a scenario is so far-fetched as
to be possible in only the barest sense, such that the only way to consider
it possible is to regard it as non-contradictory from a logical point of view.
Aristotle seems not to consider possibilities of this sort as genuine. It will be
the Megarian School that makes logical possibility a standard modality to bear
in mind in philosophical theorizing.

Synchronic possibility
Maria gave the kulfi ice cream to Luca, but the ice cream spoiled. The possi-
bility of the ice cream not spoiling thus remained non-actualized. From the
point of view of many contemporary theories, non-actualized possibilities
are still genuine possibilities; they comprise a large part of what we shall
call “mere possibilities” (discussed further in Chapter 2). Aristotle seems to
have little interest in mere possibilities, and perhaps he regarded them as
non-genuine. It was Augustine of Hippo (AD 354–430) who brought the idea
of synchronic possibility into Western philosophical discourse.

Conditionals
We have seen that Aristotle discusses conditional necessity. Also, some
of the notions of possibility he entertains involve conditionals; for instance,
a power is assigned to an entity based on the fact that, if the entity is in
certain circumstances, then some other circumstances obtain (e.g. if a viper
bites Maria in the absence of an antidote, then Maria will die). Nonetheless,
Aristotle did not separately discuss conditional statements, and did not
attempt to systematically analyze them. It was again the Megarian School,
as we shall see, that first addressed the problem of assigning truth values to
conditional sentences.

9781472524263_txt_print.indd 29 01/12/2015 09:44

 30 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Future contingents
Through the notion of diachronic possibility—a notion of possibility that is
entrenched with the perspective of temporal change—we can formulate
the so-called problem of future contingents. In general, the problem deals
with the question of whether there is an open future and whether, as time
goes by, the number of options before us changes. Aristotle introduces the
problem in the De Interpretatione 9, where we also find the famous example
of a sea battle:

Let me illustrate. A sea-fight must either take place tomorrow or not, but it
is not necessary that it should take place tomorrow, neither is it necessary
that it should not take place, yet it is necessary that it either should or
should not take place tomorrow. Since propositions correspond with facts,
it is evident that when in future events there is a real alternative, and a
potentiality in contrary directions, the corresponding affirmation and denial
have the same character.
—Trans. E. M. Edghill; available at The Internet Classics Archive: http://
classics.mit.edu/Aristotle/interpretation.1.1.html

The problem of future contingents relies upon the idea—intuitive during

Aristotle’s time and until modernity—that sentences can change in their
truth value across time; so, our sentences regarding tomorrow’s possible sea
battle may change value as time goes by. The problem involves three crucial
principles of logic, and shows that, in the case of future contingent proposi-
tions, these principles do not square with one another. The principles are:

(B) Bivalence: Every sentence expressing a proposition is either true or

false.
(EM) Excluded Middle: For every sentence S expressing a proposition,
“S or non-S” is always true. (Hence, there is no intermediate truth-
value for sentences.)
(NC) Non-Contradiction: For every sentence S expressing a proposition,
it is always true that: “It is not the case that: S and non-S” is true.
(Hence, it can never be the case that, at a same time, both S and
non-S are true.)

In the case of the sea battle example, we intuitively accept that it is true that
“either there will be a sea battle tomorrow or there won’t be one” (so, the
excluded middle holds); but, we intuitively deny that it is true that there will be
a sea battle tomorrow, and we deny that it is true that there will not be a sea
battle tomorrow (so, bivalence is given up); at the same time, we intuitively

9781472524263_txt_print.indd 30 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 31

deny that both sentences are true (so, non-contradiction holds). However,
denying bivalence cannot easily be squared with maintaining the excluded
middle and non-contradiction, hence the quest for alternative explanations.
The problem of future contingents has occupied scholars for centuries
and still, today, steers much controversial discussions (see Hasle and
Øhrstrøm (2011)). It is also in an attempt to provide a clear framework for
dealing with this problem that so-called temporal logics were developed in
the late 1950s (see Prior (1957)); temporal logics influenced to a great extent
the development of modal logic in the late 1950s and the early 1960s. We
cannot examine the discussion of the problem in detail, but members of
the Megarian School, who came up with a more refined analysis of future
contingents, put forward what is known as the Master Argument, which shall
occupy us in the next section.

§1.1.3. The Megarian School

The Megarian School (sometimes referred to as the Megaric School) was a
philosophical movement first established by Euclides of Megara (c. 435–365
BCE; Megara, Greece), and most active between the fourth and third
century BCE. Some of the members of the School—sometimes referred to
as “dialecticians”—distinguished themselves by their work in logic, which
was particularly fruitful in matters related to modalities. We shall focus on
two figures who deserve special attention for having initiated some topics
of philosophical conversation that are still alive today: Diodorus Cronus (died
c. 284 BCE), and his student, Philo, for whom we have scant biographical infor-
mation. While today we recognize the originality and importance of Diodorus
and Philo, only fragments of their writings have survived; most of what we
know of their ideas is due to later Roman philosophers, such as Sextus
Empiricus (AD 160–210), and also Diogenes Laertius (third century AD), who
wrote about them centuries after their death. Another key figure for the trans-
mission of Diodorus’s and Philo’s ideas to Medieval scholars, and, therefore,
all subsequent philosophers, is Severinus Boethius (AD 480–524); Boethius’s
work on logic and modalities is vast and articulated and deserves a separate
treatment. Here, however, we shall confine ourselves to three central contri-
butions of Diodorus and Philo, concerning conditionals, modalities, and the
so-called Master Argument.

Conditionals
Philosophers within the Megarian School seem to have been the first among
Ancient philosophers to accord a central role to propositions. Aristotelian

9781472524263_txt_print.indd 31 01/12/2015 09:44

 32 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

logic took terms (in contemporary parlance: subject and predicate) as the key
ingredients of logical analysis; for Diodorus and Philo, the fundamental units
are instead axiómatas, which roughly correspond to what we now call propo-
sitions. These stand for the content of a sentence, so that different sentences
can express the same proposition: for example, when we say that “It rains”
and Piove convey the same content—that it is raining—in two languages,
English and Italian. Unlike most contemporary authors working on proposi-
tions, however, Diodorus and Philo accept that propositions can change their
truth value over time, so that the content of a sentence can be—say—true
today, and yet false tomorrow.
The analyses of conditionals proposed by Diodorus and Philo crucially
relied on their usage of propositions and inspired contemporary treatment
of the issue. According to them (and to us), a conditional is a complex
proposition formed by two simpler propositions—called respectively the
antecedent and the consequent—that are related by an “if …, then …”
relation. For instance, the sentence “If it rains, then I go surfing” expresses a
complex proposition formed by the two propositions It rains and I go surfing,
which are related by the “if …, then …” relation. By conceiving of conditionals
in this way, Diodorus and Philo could appreciate their logical complexity and
point out difficult questions that are still open for us. The central question
being: When is it that a conditional proposition is true?
Conditionals are tricky because, in some instances, while we may know
the truth conditions of each of the propositions composing a conditional,
we may still not know whether the conditional is true or false. For instance,
consider the sentence: “If it rains, then in one hour there are sixty minutes.”
The consequent of the conditional is true, since by definition there are sixty
minutes in one hour; but is the conditional as a whole true? No matter
whether it is raining or not, the problem seems to be that the antecedent and
the consequent are not relevantly related.
Diodorus and Philo gave two different solutions to this problem.
Diodorus’s analysis was more complex, taking into consideration the time
at which a proposition is true or false: he seems to have maintained that
a conditional is true if it is not possible, and it was not possible, that its
antecedent is true and its consequent false. By specifying the time at which
a proposition is true, Diodorus aimed to devise stricter truth conditions for
conditionals; the texts we have that recount Diodorus’s position, however,
do not offer sufficiently clear information to allow us to reconstruct its
subtleties in a concise and uncontroversial manner. The solution he is
aiming for, in any case, seems unable to address the problem of relevance
expressed in the previous paragraph: for Diodorus the sentence “If it rains,
then in one hour there are sixty minutes” seems to have expressed a true
proposition.

9781472524263_txt_print.indd 32 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 33

Philo, on the other hand, proposed an analysis of conditionals that had

much good fortune and is still endorsed in contemporary classical logic.
According to Philo, a conditional is false just in case its antecedent is true and
its consequent false; otherwise, it is true. Thus, for Philo, too, the sentence
“If it rains, then in one hour there are sixty minutes” expresses a true propo-
sition. But Philo’s solution is straightforward: the truth of a conditional is
evaluated in a simple way by taking into account the truth conditions of its
antecedent and its consequent.
Conditionals form an important part of the contemporary debate on the
metaphysics of modality. It is by means of conditionals that we express
the possibility of many scenarios. To name just one notable instance, it is
by means of a kind of conditional—so-called counterfactual conditional or
counterfactual—that we express the possibility that matters could have
been otherwise, as when we say that Napoleon could have won at Waterloo.
Counterfactuals are quite relevant to the purposes of this volume: for
example, they play a crucial role in our understanding of scientific laws, a
central topic in the metaphysics of modality; and David Lewis, a key figure in
the contemporary debates on the metaphysics of modality, offered the first
articulated account of his modal realism (which we will take up in Chapter 3)
in a book that dealt with counterfactuals (Lewis 1973).

Possibility and necessity in Diodorus and Philo

We cannot pass by the work that Diodorus Cronus and Philo did on modalities,
though we won’t have the space to analyze it. Again, an element of originality
of their analyses is that they are carried forward in terms of propositions. But,
most importantly, they provided us with two definitions of possibility that we
do not find in Aristotle. (i) According to Diodorus, a proposition is possible just
in case it is true now or it will be true in the future. This is a strict notion of
possibility, which leaves no room for mere possibilities. Diodorus also offered
a parallel notion of necessity that is also bound to time constraints: a propo-
sition is necessary just in case it is true now and it will never be false in the
future. (ii) Philo seems rather to have endorsed the logical notion of possibility
as the absence of contradiction: a proposition is possible just in case, by its
own nature, it is capable of being true. The expression “by its own nature”
suggests that Philo is thinking that it is only a self-contradictory element that
can render a proposition not possible. As we saw, the logical notion of possi-
bility was missing in Aristotle. The merit of Philo’s work is not only to have
introduced logical possibility, but to have clearly defined necessity in terms of
possibility: a scenario is necessary just in case it is not possible that it is not
the case; and vice versa: a scenario is possible just in case it is not necessary
that it is not the case; and he did all of this with propositions.

9781472524263_txt_print.indd 33 01/12/2015 09:44

 34 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

The influence of Diodorus Cronus and Philo on the contemporary devel-

opment of modal logic and modal metaphysics can hardly be overestimated.
Additional important contributions from Ancient philosophy were made by
the Stoics: in particular, by Chrysippus (c. 279–206 BCE, Athens), who defined
possibility in terms of the absence of external conditions of impediment. But
it is especially through the works of Boethius that the legacy of Megarian and
Stoic philosophers has reached us (see Bobzien (2011)).

The Master Argument

We have seen that Diodorus’s work on modalities is particularly sensitive to
the logical and philosophical conundrums of time: the definitions of modal
notions he offered all take into account the fact that propositions can change
their truth value over time. It is, therefore, not surprising that he put forward
one of the most formidable problems concerning the philosophy of time,
which is closely connected to the problem of future contingents discussed
by Aristotle. Diodorus’s problem can be expressed through the so-called
Master Argument. In its simplest formulation, the argument asserts that the
following three claims cannot all be true:

(M1) Every true proposition that is about the past is necessary.

(M2) A proposition that is possibly true cannot imply a proposition that is
necessarily false.
(M3) There is a proposition that is possible, but that neither is nor will be
true.

Diodorus argued that the problem rested in (M3). As a reader may recognize,
in fact, (M3) violates Diodorus’s definition possibility, according to which a
proposition is possible just in case it is true now or it will be true in the future.
This definition is highly controversial, however, because it suggests a strict
notion of possibility. Alternatively, we could maintain that (M3) is true, but
give up either (M1) or (M2). But since both (M1) and (M2) make substantial
claims about propositions that are also superficially plausible, it becomes
tricky to understand which of the two claims should be discharged.
Because of its centrality, the Master Argument has had several alter-
native formulations: most notable are the ones by Prior (1957), Lavenham
(see Hasle and Øhrstrøm (2011)), and Gaskin (1995). During Medieval
times, the argument gained centrality in connection with the problem of
divine foreknowledge: if God knows the future—and thus facts about the
future seem to be already determinately true or determinately false—is
the future contingent or do we simply have an illusion of contingency? For
instance, if God knows whether there will be a sea battle tomorrow, is it

9781472524263_txt_print.indd 34 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 35

contingent that there will be a sea battle tomorrow? If facts about the future
are not contingent, then are humans really free agents who are responsible
for their actions?

§1.2. Modalities in Medieval philosophy

§1.2.1. The Arabic and Scholastic traditions
It is from the ideas of the Ancients—such as Aristotle, Diodorus, and Philo—
that much of the contemporary debate in modal metaphysics takes root.
But it would be an error to overlook the fact that those ideas were selected,
transcribed, and often reformulated by philosophers that came later, and
that, little by little, gave shape to the Western philosophical tradition. Today
we read and discuss the ideas of Aristotle, Diodorus, and Philo through the
lenses set by those philosophers. Among them, we find eminent figures of
Late Antiquity, such as Boethius (c. 480–524, Rome). And we also find several
Arabic philosophers who made a decisive contribution to the development
of modal logic and metaphysics, such as Al-Fārābī (c. 872–950, Baghdad),
Avicenna (c. 980–1037, Persia), Al-Ghazali (1058–1111, Khorasan, Iran),
Averroes (1126–98, Córdoba), and Kātibī (died 1276, Persia). Finally, we find
those philosophers who operated in the new university system that emerged
throughout Europe in the eleventh and twelfth centuries, and that formed
the so-called Scholastic tradition; this group includes figures such as Abelard
(1079–1142, Paris), Kilwarby (c. 1215–79, Paris and Oxford), Aquinas (1225–74,
Montecassino, Italy, and Paris), Scotus (c. 1266–1308, Oxford and Paris), and
Gersonides (1288–1344, Avignon).
We cannot examine the details of the history of modal logic and metaphysics
during this entire period; the reader will find ample suggestions for how to do
so in the Further Reading section. Nonetheless, to convey a small sense of
the relevance of the figures just listed, we shall point out just the innovative
elements of two authors. One is Avicenna: he developed both a modal syllo-
gistic (in the style of Aristotle) and a modal predicate logic (in the style of the
Megarian School); most importantly, he understood each term, predicate,
or proposition as standing for all of its possible applications, thus setting
up a much more powerful theoretical framework than the one provided by
the Ancient philosophers. The second author is Scotus, who forged a modal
theory based on logical possibility and compossibility: a scenario is possible
just in case existence is not in contradiction with it, though the scenario may
not be compossible with others. Scotus’s view comes close to the standard
way of thinking about possibility in contemporary work. It also had a major

9781472524263_txt_print.indd 35 01/12/2015 09:44

 36 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

influence over two central figures in our historical journey: Ockham and
Buridan. Finally, we shall also remark that Arabic and Medieval authors greatly
developed and enriched the Aristotelian theory of substances, including the
associated doctrine of essentialism which will be discussed at various points
in the remaining chapters (especially 3, 7, and 8).
Certain important ideas that characterize the contemporary debates on
modalities in logic, semantics, and metaphysics have their origins in (and
often were anticipated by) the works of William Ockham (c. 1287–1347,
London and Munich) and John Buridan (c. 1295–1358, Paris). Needless to
say, the ideas of these two philosophers developed in conversation and
disagreement with their contemporaries. Still, because the views of Ockham
and Buridan dominated the debates on modality in the fourteenth century, it
is instructive to examine them.

§1.2.2. Ockham
Ockham was a prominent figure in England and continental Europe during
the first half of the fourteenth century. He is best known for having defended
a sophisticated and influential version of nominalism, which got him into
trouble with the official representatives of the Church at the time (Ockham
was a Franciscan friar, and, in the 1320s, he was called to defend his views
both before his Franciscan province chapter in Bristol and before the Pope in
Avignon). Looking at Ockham’s nominalism will help us understand his contri-
bution to modal logic and metaphysics. Nominalism is the metaphysical view
according to which all there is to the world are particular entities. It stands
in opposition to realism, the metaphysical view that there are both particular
and universal entities. Consider for instance the sentence: “This one and that
one are blue.” For the Realist, the sentence is about two particular individuals
(“this one” and “that one”) and a universal (“blue”), which is common to both
particulars; this strikes us as consistent with our most intuitive worldview.
The Nominalist, instead, faces the challenge of explaining the truth of this
sentence by means of particulars alone.
In order to deal with sentences such as “This one and that one are
blue”, Ockham devised an eliminativist strategy, which can be summarized
in two steps (see also Read (forthcoming)). First, deny that universals are
real; in our case, deny that “blue” is anything more than the creation of a
human mind, so that when one attributes the color blue to some particular
individual, one is simply describing a projection of one’s mind. Second, deny
that there is any object of thought—what some today call “concepts”—that
is universal. When we use the word “blue” we are not employing the same
concept over and over; rather, we are having a particular thought, which

9781472524263_txt_print.indd 36 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 37

happens to be appropriately related to previous thoughts and uses. Each

time we use the word “blue” to describe a particular, it is as if we are intro-
ducing a new name for that particular: hence the label “nominalism”. The
Medieval debate about nominalism and realism can readily be related to
contemporary disputes in philosophy of mind, metaphysics, and semantics,
standing as an example of Ockham’s continuing significance in philosophy.
Indeed, that very debate still occupies contemporary discussions in those
fields.
The strategy Ockham embraced to argue for nominalism is recognizable
also in the specific contribution he made to the analysis of modalities. First of
all, Ockham pursues his analysis in terms of so-called propositional dictums,
which correspond to the English “that” clause (e.g. “That Ockham was
smart, everybody knows!”) Second, Ockham distinguishes between two
types of modal propositions: those having a divided sense and those having
a composite sense. This distinction roughly corresponds to the contemporary
distinction between de re and de dicto, which we shall discuss further in
Chapter 2. Here are two examples:

Divided Modality: “Every philosopher necessarily likes Ockham.”

Composite Modality: “Necessarily: every philosopher likes Ockham.”

And here is how the two examples are analyzed in terms of Ockham’s
nominalist strategy:

Divided Modality: Every proposition where “likes Ockham”

is predicated of any specific philosopher is
necessary.
Composite Modality: Every proposition “Every philosopher likes
Ockham” is necessary.

Notice that the analyses refer to every proposition of a certain sort. This
is because there is no single, universal proposition serving as the content
of different sentences or utterances of the same sentence: that is, “Every
philosopher likes Ockham” stands for a different proposition on every
occasion on which it is uttered or entertained. The key point in Ockham’s
strategy is to reduce all modalities to attitudes that an agent takes toward
propositions; the necessary (like any modality) is simply a way human beings
regard propositions. For Ockham, indeed, there is no necessary being
but God. Ockham’s metaphysics of modalities is thus minimal, but such
minimality is the outcome of a sophisticated semantics.
Our discussion of Ockham’s analyses of modality should suffice to demon-
strate how sophisticated the logical, semantic, and metaphysical theories of

9781472524263_txt_print.indd 37 01/12/2015 09:44

 38 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

modality had become by the fourteenth century. With Buridan, we reach the
most mature point of Medieval logic and metaphysics.

§1.2.3. Buridan
Buridan lived, studied, and taught in Paris. Like Ockham, he also defended
nominalism; but, unlike Ockham, he did not get into trouble with the Church.
Buridan’s metaphysics is oriented by his views in semantics and logic. His
theory of meaning was radical and ingenious. To start, he maintained that
any term stands for a certain particular entity regardless of whether the
proposition in which the terms appear is true or false. Hence, according to
Buridan, the truth of a proposition does not depend on whether the terms
in the proposition stand for (or pick out) something; rather, a proposition is
true when the terms stand for entities that are in the appropriate relation. For
instance, the sentence “This one is blue” is true when “This one” stands for
the same entity as “blue”. (Presumably, the sentence “This one and that one
are blue” would be analyzed in terms of distinct identity statements, “This
one is blue” and “That one is blue”.)
With respect to propositions involving modalities, Buridan inherits the
general strategy of Ockham, which used “that” clauses and distinguished
between divided and composite modalities. But Buridan improved Ockham’s
approach in at least two respects. A first improvement is quantifying directly
over “that” clauses, in the analysis of modal propositions. As such, the
following two examples:

Divided Modality: “Every philosopher necessarily likes Buridan”;

Composite Modality: “Necessarily: every philosopher likes Buridan”

… are analyzed by Buridan as follows:

Divided Modality: Every that-likes-Buridan, when predicated of any

specific philosopher, is necessary.
Composite Modality: Every that-every-philosopher-likes-Ockham is
necessary.

The second improvement is distinguishing three senses in which a propo-

sition may be said to be necessary: conditional, temporal, and absolute.
This move allows Buridan to concede the necessity of some propositions
without giving up his nominalist conviction that there are no universal laws,
no universal ties, but only particular relations.
Consider the sentence:

9781472524263_txt_print.indd 38 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 39

(W) “Necessarily: wisdom is a virtue.”

Sentences such as (W) are problematic for the nominalist because they seem
to be about universals, such as wisdom and virtue; thus, it is not clear how
the nominalist may accept their truth without thereby affirming that there are
universals. Instead of taking (W) at face value, Buridan maintains that (W) can
be interpreted as making one of the three following claims:

Conditional: “If there is something that is wise, then that thing is virtuous.”
Temporal: “If at some time there is wisdom, then it is a virtue.”
Absolute: It is not possible that at any time: “Wisdom is not a virtue.”

The conditional and temporal readings of (W) commit only to the existence of
particulars—a particular entity that is wise, and a particular temporal instance
of wisdom. And all it takes to falsify the third reading of (W) is an instance of
a particular that happens to be wise and not virtuous.
Buridan and Ockham illustrate, perhaps to an extreme, a methodology that
is still much in vogue among analytic metaphysicians: they settle on a certain
metaphysics (nominalism), but do not employ metaphysical arguments to
defend it, appealing to sophisticated and subtle semantic machinery instead.
As we shall see, such a strategy is common in the contemporary debate on
the metaphysics of modality. Contemporary scholars still have much to dig up
and learn from the writings of the Arabic and Scholastic authors.

§1.3. Modalities in Early Modern philosophy

A few decades after Buridan’s death, in the early 1360s, European society started
a process of deep transformation, which led also to a change in how culture
was produced and disseminated, and eventually resulted in Humanism and the
Renaissance. By the early seventeenth century, European society, and European
philosophy with it, was facing entirely new questions, and many protagonists
of the philosophical debate were not academics, but rather laypeople. The logic
and semantics of modalities were no longer of much concern; epistemology,
metaphysics, and ethics took center stage. And yet the Early Modern period
is especially rich in discussion of modalities. The most representative texts
also happen to be quite accessible. For this reason, in what follows we shall
exemplify some key ideas through the voices of their authors. Though the list of
important figures is long, we will limit our survey to four: Descartes (1596–1650,
France, Holland, and Sweden), Spinoza (1632–77, Holland), Leibniz (1646–1716,
Germany and Central Europe), and Hume (1711–78, England and France).

9781472524263_txt_print.indd 39 01/12/2015 09:44

 40 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

§1.3.1. Descartes: Thought experiments and

God’s omnipotence
Descartes is canonically regarded as one of the first major Early Modern
philosophers. Two aspects of his work contribute most distinctly to the
debates on the metaphysics and epistemology of modality: his radical
skeptical challenge and his conception of God’s omnipotence. (A third
important aspect is his analysis of free will, but since this volume will not
directly tackle this issue, we shall leave it aside.)
At the end of the first of his Meditations, Descartes advances the
hypothesis that there is an evil deceiver that is as powerful and ingenious
as God, and who employs all its energies to deceive him with respect to
what is true and false. This hypothesis is then used to show that one kind of
self-knowledge (the so-called cogito) is infallible, and is, from an epistemic
point of view, foundational; the centrality of self-knowledge, furthermore, is
employed by Descartes to establish his metaphysics of God, the person, and
every other substance.
From our perspective, the remarkable aspect of Descartes’s argumentative
strategy is the use of a thought experiment, where a seemingly impossible
scenario is used to provide a fresh start in epistemology and metaphysics.
Such a strategy did not go unnoticed by Descartes’s contemporaries and
remains controversial. Not only did Descartes have to defend himself from
the charge of believing in the existence of an evil god, or that God is evil; more
interestingly for us, Descartes had to justify his method: in what way can we
gain philosophical insight from considering not simply a scenario that is false,
but a scenario that is in all likelihood impossible? Here is what Descartes had
to say in response:

A philosopher would be no more surprised at such suppositions of falsity

than he would be if, in order to straighten out a curved stick, we bent it
round in the opposite direction. The philosopher knows that it is often
useful to assume falsehoods instead of truths in this way in order to shed
light on the truth.
—Fifth Replies, AT 7:349–50, CSM 2:242

Descartes’s reply is quite important to the theme of our book. He is indeed

defending a strategy that is by now recognized as standard in many sciences.
Not only do models and principles (in the natural, social, and humanistic
sciences) typically rest on ideal scenarios; they often employ impossible or
unlikely scenarios.
From an epistemological point of view, therefore, we seem to be justified

9781472524263_txt_print.indd 40 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 41

in entertaining thoughts about the impossible; but what is the content of such
thoughts? Do such impossible scenarios exist and, if so, in what sense? This
is the root of the concern that, by using the hypothesis of the evil deceiver,
Descartes was committing himself to the existence of an evil god. We shall
discuss these and related questions at the end of Chapter 8.
The second aspect of Descartes’s philosophy that deserves to be highlighted
in this context is his conception of God’s omnipotence. Descartes develops in
an original way the thesis that God is the only necessary being. Since God
is the only necessary being and God is omnipotent, then any other truth is
contingent. Thus, Descartes maintains that ostensibly necessary truths, such
as mathematical or geometrical truths, are contingent upon God’s will:

And even if God has willed that some truths should be necessary, this
does not mean that he willed them necessarily; for it is one thing to will
that they be necessary, and quite another to will this necessarily, or to be
necessitated to will it.

In order to grasp the meaning of the position defended by Descartes we

arguably must help ourselves—as he does, in the case of the evil deceiver—to
certain impossible scenarios. To conceive of the contingency of mathematical
or geometrical truths implies the ability to conceive of impossible scenarios.
But can we really do that? Again, this is a question that runs deep into the
contemporary debates on the metaphysics of modality and that Descartes’s
work vividly brings to light.

§1.3.2. Spinoza’s necessitarianism

Spinoza’s work is closely entangled with the metaphysics of modality and
contains some radical and original views. We shall encounter it again in
Chapter 2, when examining so-called modal expressivism. Here, we will
confine ourselves to Spinoza’s necessitarianism, a position that we encoun-
tered earlier when discussing the ideas of Parmenides and Zeno. As we
stated at that point, necessitarianism rests on the acceptance of the two
following claims: (i) the universe exists out of necessity; (ii) the universe is
necessarily the way it is. Spinoza seems to have accepted both. Indeed, for
him, God exists out of necessity, and God is necessarily the way God is. If to
this we add that God is the only possible substance, then we have the result
that God is the universe, which exists out of necessity and is necessarily the
way it is.
Spinoza defended his necessitarianism by providing ample metaphysical
details of his position. His view seems very different from the ones of

9781472524263_txt_print.indd 41 01/12/2015 09:44

 42 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Parmenides and Zeno. The latter seem to have rejected all superficial impres-
sions suggesting the existence of a multiplicity of entities; Spinoza, instead,
has a place for every entity, including ordinary ones: they are all in God, as
modes of existence of some of God’s attributes (extension and thought).
For instance, consider the naïve belief that there are a cup and a cell phone
on the table. While Parmenides and Zeno would dismiss this belief as false
because it postulates the existence of three entities, Spinoza would agree
that there are three entities, but they are all modes of existence of one of
God’s attributes—extension.
Necessitarianism implies determinism, the view that no particular event
could have been otherwise. In order to defend his view, Spinoza also puts
forward an original account of causation as conceptual entailment. If God is
the cause of the fact that the cup and the cell phone are on the table, this
is because it follows from the concept of God that the cup and the cell phone
are on the table. We shall see that Hume’s position with respect to causation
is at the other end of the spectrum with respect to Spinoza’s: for Hume, the
concept of the cup is in no way related to any other concept and, indeed,
causation is not a real phenomenon. There is no need to enter into the details
of Spinoza’s view here; it suffices to point out that, in the contemporary
debate on the modal status of scientific laws, Spinoza offers an original and
important perspective.

§1.3.3. Leibniz: Possible worlds, theodicy,

and super-essentialism
Leibniz is a central figure in the history of modality, if for no other reason than
that he advanced a theory of modality in terms of possible worlds that greatly
influenced the development of possible-worlds semantics in the twentieth
century (see Chapters 2 and 3, and Menzel (2015)). Also connected to possible
worlds is a second aspect of Leibniz’s thought that we shall stress, that is, his
theodicy, according to which we live in the best of all possible worlds. Finally,
the third aspect of Leibniz’s work that deserves to be examined is his doctrine
of so-called super-essentialism. Let us look at each in turn.

§1.3.3.1. Possible worlds

Leibniz’s views in metaphysics are presented over a vast array of writings;
they are also tightly interwoven with each other, and the texts raise no
small interpretive difficulties. We can provide not even an overview of the
big picture here; more modestly, we shall limit ourselves to explaining the
role of possible worlds in Leibniz’s metaphysics. Leibniz may not have been

9781472524263_txt_print.indd 42 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 43

the first philosopher to employ the notion of a possible world; however, he

certainly is the first to have established it as central to philosophical discus-
sions of modality. He seems to have begun using the expression “possible
world” in the 1680s, using the idea to explain how his notion of a substance
(e.g. a specific person or a specific organism) left room for contingency. The
issue was not a minor one. For Leibniz, a substance is a complete concept:
that is, at any given time, a substance is so determined that it contains all the
properties that are true of it. So, while Leibniz was still alive, it was already
true that this very book discusses Leibniz, and the complete concept of
Leibniz contained this property (viz. the property that there is a book partly
about Leibniz). In a famous letter to Arnauld of July 14, 1686, Leibniz explains
how his metaphysics leaves room for contingency by appealing to possible
worlds:

I think there is an infinity of possible ways in which to create the world,

according to the different designs which God could form, and that each
possible world depends on certain principal designs or purposes of God
which are distinctive of it, that is, certain primary free decrees (conceived
sub ratione possibilitatis) or certain laws of the general order of this
possible universe with which they are in accord and whose concept they
determine, as they do also the concepts of all the individual substances
which must enter into this same universe.
—G II 51/L 333

One way to read Leibniz, which was made prominent by authors such as
Carnap (1947), Mates (1972), and Adams (1974), is that possible worlds are
ideas entertained by God. Each such idea is specified in all its details; it is
a complete and consistent story about how the world could have been. In
this sense, every world is fully determined. In the actual world, for example,
Leibniz’s destiny was already determined: this book had to discuss him;
nonetheless, there is another world, which may be very similar to ours, where
this book does not touch on Leibniz. Now, God chose to create the actual
world over any of the others, but he might have done otherwise. The act of
creation was the “law of the general order”, by means of which this world—
determined in all its details—came to be.
Leibniz’s theory of possible worlds suggested a simple and intuitive
analysis of modality that had not been seen before. According to it, a propo-
sition is possible just in case it is true in at least one possible world; it is
contingent just in case it is true in at least one possible world and false in at
least one possible world; and it is necessary when it is true in all worlds.
It is unclear whether Leibniz’s metaphysics is indeed compatible with the
picture of possible worlds as complete and consistent stories (see Rudder

9781472524263_txt_print.indd 43 01/12/2015 09:44

 44 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Beker (1985) for a criticism of such an interpretation). At any rate, Leibniz’s

suggestion has had a long career in the history of modality, and is of central
importance today: it is at the core of the most popular modal semantics and
of several alternative metaphysical theories of modality; moreover, the idea of
possible worlds as complete and consistent stories is today a view called
linguistic ersatzism (see Chapter 5).

§1.3.3.2. Theodicy
Possible worlds are employed by Leibniz also to answer a daunting question
in ethics and theology: if God is infinitely powerful and infinitely good, how is
it possible that there is evil in the world for which no human can be blamed?
For instance, why is it that some people suffer—say, from random medical
illnesses or geological disasters—without anyone being responsible for their
suffering? Any attempt to answer this question is called a theodicy: that is, it
is a justification of God’s attributes in light of such forms of evil.
Leibniz’s solution is that God created the best of all possible worlds. There
is no “innocent suffering”, no gratuitous evil that takes place in our world,
although this is not how things appear to us. We may fail to appreciate
how any instance of evil is connected to some good because of the infinite
complexity of the universe. But, God, who understands the complexity,
chose for the best. In the Candide, Voltaire ridicules Leibniz’s solution. Doubts
about its soundness notwithstanding, the solution is an excellent example
of the relationship between the metaphysics of modality, ethics, and the
philosophy of religion.

§1.3.3.3. Super-essentialism
Before moving on, it is important to mention another notion that comes out of
Leibniz’s philosophy and that is of importance to contemporary metaphysics
of modality: the so-called super-essentialism. We shall return to essentialism
in Chapters 3, 7, and 8 but, in brief, an essence is any property that belongs
to an individual unconditionally (at all times and in all possible scenarios).
Some analyze such unconditionality in modal terms, suggesting that essential
properties belong to an individual out of (metaphysical or natural) necessity;
others, instead, analyze the unconditionality in terms of identity: an essence
is a property that defines an individual. Leibniz’s theory of substances—
according to which substances are complete concepts—suggests a radical
form of essentialism, super-essentialism: all the properties of an individual
belong to it unconditionally. That is, at all times and in all possible scenarios,
an individual (a substance, in Leibniz’s terminology) never changes its
properties.

9781472524263_txt_print.indd 44 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 45

§1.3.4. Hume on separability

The fourth and last Early Modern author we shall discuss is Hume. His
views on possibility and necessity stand in stark contrast with those of
Spinoza, and have greatly influenced contemporary discussions on the
epistemology and the metaphysics of modality. For instance, it is difficult to
understand the work of a prominent contemporary author in the field, David
Lewis, without reading Hume’s work. Moreover, Hume’s views on possibility
and necessity are crucial to appreciating Kant’s metaphysics.
The key modal notion for Hume is separability. For him, each individual
is separable from any other, and so is each event. While Leibniz maintained
that the concept of an individual was so rich as to contain all of its properties,
Hume strips down the concept of an individual to its bare minimum—that
which we perceive of it:

From the first appearance of an object, we can never conjecture what

effect will result from it. But were the power or energy of any cause
discoverable by the mind, we could foresee the effect, even without
experience; and might, at first, pronounce with certainty concerning it, by
the mere dint of thought and reasoning.
—EHU 7.1.7; SBN 63

Since individual entities and events are separable, there is no contradiction in

thinking that any individual (or any event) can be combined in any way with
any other (as long as, indeed, no contradiction ensues). Thus, there is no
necessary connection in nature between individuals, which may be justified
on the basis of our definition of said individuals:

It has been observ’d already, that in no single instance the ultimate connexion
of any objects is discoverable, either by our senses or reason, and that we
can never penetrate so far into the essence and construction of bodies, as
to perceive the principle on which their mutual influence depends.
—T 2.3.1.4; SBN 400

As we have seen, the notion of the possible as that which is non-contra-

dictory can be traced back to Philo, and it was not until Scotus that this
notion received more systematic treatment. However, Scotus was a realist,
who worked with a rich notion of the individual (or substance). So, Hume is
typically a “modal hero” to the empiricists working with the notion of possi-
bility as non-contradiction. His legacy, furthermore, includes another really
important idea: that conceivability is a reliable guide to possibility:

9781472524263_txt_print.indd 45 01/12/2015 09:44

 46 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

The mind can always conceive any effect to follow from any cause, and
indeed any event to follow upon another: Whatever we can conceive is
possible, at least in a metaphysical sense.
—T Abstract 11; SBN 650

Hume’s ability to couch original ideas about the metaphysics of modality in

a simple and direct style is exemplary. Whether he is right is still a widely
debated issue, which we shall confront a few times in the rest of the volume
(for instance, Chapters 3 and 6).

§1.4. Towards quantified modal logic and

possible-worlds semantics
It would be unfair to say that modalities were not studied in nineteenth-
century Western philosophy; we find them discussed, for instance, in such
masterpieces as Hegel’s Science of Logic (1812–17) and Mill’s A System of
Logic (1843). Yet, modalities did not take center-stage in this period either: key
disputes do not center around the understanding of modalities and there is
arguably little originality in the ways they are understood.
Still, nineteenth-century developments in logic quietly sowed the seeds
of a new era in the study of modalities. The innovations of Mill’s A System
of Logic and Boole’s An Investigation of the Laws of Thought (1854) culmi-
nated in the seminal works of Frege and Russell, preparing the ground for a
new logic and semantics of modality. It would be C. I. Lewis (1918) who first
developed Quantified Modal Logic, in connection with which possible-worlds
semantics made its appearance at the end of the 1950s. The details of this
story will be discussed in Chapters 2 and 3.
The study of modalities over the history of Western philosophy has neither
been guided by a fixed set of questions nor required a fixed set of philo-
sophical skills. As we have seen, modalities have been approached from the
viewpoints of logic, semantics, metaphysics, ethics, and epistemology. Also,
because modalities are connected with so many branches of philosophy,
their study can be particularly fruitful. In the rest of this volume, we shall
concentrate on the metaphysical details of the contemporary debate about
modalities, leaving aside the difficulties confronting those who want to
provide an adequate logical or semantic treatment. Nonetheless, questions
of semantics, logic, and epistemology will surface from time to time, and will
accompany us in our journey.

9781472524263_txt_print.indd 46 01/12/2015 09:44

 THE METAPHYSICS OF MODALITY: A HISTORICAL OVERVIEW 47

Study Questions
• What is Parmenides’s metaphysical monism?
• What is necessitarianism?
• What was the goal of Zeno’s paradoxes?
• What are Zeno’s four paradoxes of motion?
• What are Zeno’s two paradoxes of plurality?
• What is Zeno’s millet paradox?
• What is Zeno’s paradox of place?
• How does Aristotle define general necessity?
• What is the distinction between conditional and absolute necessity in
Aristotle?
• What are the three varieties of conditional necessity, according to Aristotle?
• What are the three senses of possibility introduced by Aristotle?
• What are the three senses of possibility that Aristotle fails to sort out?
• What is the problem of future contingents?
• What are the three main contributions of the Megarian School to modal
discourse?
• What is the Master Argument?
• What is the distinction between divided and composite modalities?
• What are the three senses of necessity discussed by Buridan?
• What is the controversial modal aspect of Descartes’s thought experiments?
• What are the elements of originality in Descartes’s conception of God’s
omnipotence?
• What is Spinoza’s necessitarianism?
• What is a possible world, according to Leibniz?
• What is Leibniz’s theodicy?
• What is Leibnizian super-essentialism?
• What is separability, for Hume, and why is it important for discussing
modality?

FURTHER READING
For a comprehensive overview of the main moments in the history of modal
logic, the classic by Kneale and Kneale (1962) is still a good starting point. Other
useful readings include: Knuuttila (1998); Gabbay and Woods (2004a, 2004b, 2006,
2008a, 2008b); and Bobzien (2014).
For an overview of Parmenides’s philosophy, including a discussion of his
metaphysical monism along with its implications for modality, see Palmer (2009:
Chapters 1–4, 2012a). As for the study of Zeno, good starting points are Palmer
(2009: Chapter 5, 2012b) and Huggett (2010).
▲

9781472524263_txt_print.indd 47 01/12/2015 09:44

 48 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

To place Aristotle’s view on modalities in the broader contexts of his

metaphysics and logic see, respectively, Cohen (2014) and Smith (2014). Many
are the monographic studies on Aristotle’s views on modality, including Hintikka
(1973), Waterloo (1982), van Rijen (1989), Patterson (1995), and Rini (2010). Among
the recent articles on the subject are Denyer and Makin (2000), Malink (2006), and
Marmodoro (2015).
On future contingents in Aristotle and the Megarian School, see the classic
by Gaskin (1995). For a history of the debates on future contingents in Late
Antiquity, see Mignucci (1989); for a history of the debates during the Medieval
period, see Normore (1982) and Knuuttila (2014); for a history of the problem of
future contingents in connection with divine foreknowledge, see Craig (1988); and
for an up-to-date framing of the debate, see Hasle and Øhrstrøm (2011). Prior
(1957) is a contemporary classic on the subject.
For an introduction to the key figures of the Megarian School, and their views
on modality, see Bobzien (1998, 2011).
The key moments of the Medieval debates on modality are surveyed in Dutilh
Novaes and Read (forthcoming), Lagerlund (2000), Thom (2003), and Knuuttila
(2008, 2013). For a comparison of Arabic and Medieval views on modality, see
Kukkonen (2000, 2005) and Knuuttila and Kukkonen (2011). Also useful to consult
are Rescher et al. (1974) and Dutilh Novaes (2007). On the developments of logic
in the fourteenth century, see Read (forthcoming); for an overview of Ockham’s
philosophy, see Spade (1999) and Spade and Panaccio (2011). As for Buridan,
good starting points are Klima (2009) and Zupko (2014).
For the connection between Medieval and Early Modern views on modality,
see Friedman and Nielsen (2003). For a discussion of modalities in Descartes’s
thought, see Curley (1984), Normore (1991), Bennet (1994), and Cunning (2014).
For a comparison between Descartes’s and Leibniz’s views on modality, consult
Wee (2006). For Spinoza’s view on necessity, see: Curley and Walski (1999); Miller
(2001); Garrett (2001); Koistinen (2003); Martin (2010); and Chignell; also useful is
Lin (2012). For Leibniz’s contribution to the development of the history of modality,
a good starting point is Look (2013); still useful to consult is the classic by Mates
(1986). Among the articles on the subject are: Blumenfeld (1973); Fitch (1979);
Mondadori (1985); Wilson (2000); Jauernig (2008); Lin (2012); and Newlands (2013).
For a discussion of Hume’s ideas on necessary connections, starting points are
Strawson (1989, 2000) and Craig (2002), but see also Garrett (1985) and Kail (2003).
For a discussion of Hume’s ideas on liberty and necessity, see Bricke (2011).
On the development of logic in the Modern period, see Gabbay and Woods
(2006, 2008b, 2009). For the modern origins of possible-worlds semantics and
modal logic, see Ballarin (2010) and the Further Reading section of Chapter 3.

9781472524263_txt_print.indd 48 01/12/2015 09:44

 2
Modal skepticism and
modal expressivism

I n this chapter, we consider the opinions of those who regard modalities as

expressing no clear conceptual content (modal skeptics) or no content at all
(modal expressivists). We will begin our analysis by examining what perhaps
cannot strictly be called a theory, but rather an attitude toward the topic at hand.
The attitude in question is skepticism, and we will see three versions of it: the
first is more attentive to certain difficulties in providing a semantics and logic of
modality, and Willard van Orman Quine was its most renowned proponent (§2.1.);
the second is more attentive to certain difficulties regarding the epistemology of
modality, and Peter van Inwagen was its most renowned proponent (§2.2.); and
the third is the more naïve and radical approach, but is, nonetheless, relevant
to our discussion (§2.3.). Afterward (§2.4.), we will take up modal expressivism,
according to which alethic modalities express sentiments of speakers in much
the way that the expressions “woohoo!” and “boo!” express sentiments of,
respectively, approbation and disapprobation on the part of a speaker. Modal
expressivism is an appealing approach to modalities, and it is especially useful to
test one’s own views; it suffers from two major problems, which we will address.

§2.1. Modal skepticism: Preliminary remarks

S kepticism is a philosophical attitude rooted in doubting. Philosophers

who endorse skepticism, or take its challenges very seriously, can
be found in every age and place. Among the chief figures are Arcesilaus,
Carneades, Aenesidemus, Sextus Empiricus, Michel de Montaigne, René
Descartes, David Hume, George E. Moore, Ludwig Wittgenstein, and Hilary

9781472524263_txt_print.indd 49 01/12/2015 09:44

 50 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Putnam. Skeptical doubting is peculiar, and should be distinguished from a

different form of doubting. If Giovanni tells you that today is the 6th and you
firmly believe that it is the 7th, you will doubt Giovanni’s words because you
strongly believe you have the truth otherwise. This is not skeptical doubting.
Consider instead a different scenario. Giovanni tells you he believes that
President Obama is a KGB spy; you do not have any way to disprove him,
but you are not confident that his source is reliable. In this case, you do not
believe you have the truth otherwise: more modestly, you refuse to commit
to either the truth or the falsity of Giovanni’s claim. You are now raising a
skeptical doubt towards that claim, flagged by your suspension of judgment.
Skeptical doubts can be differentiated based on their targets. For instance,
if you doubt that someone knows that such and such is the case, your doubt
goes under the umbrella of epistemic skepticism. If instead you doubt that
something exists, as when you doubt that God exists or that there is a ghost
in the house, your doubt goes under the umbrella of ontological skepticism.
The most famous formulations of skepticism are of the epistemic variety.
Particularly famous was the skeptical method developed by Arcesilaus of
Pitane, the philosopher who took the directorship of Plato’s Academy about
seventy years after his death, in 272 BCE. Arcesilaus’s skepticism, which
goes under the name of “Academic Skepticism”, applies across the board, to
any sort of target. It is so radical that it cannot even be formulated. Without
entering into details that are unnecessary here, the gist of the position is
this: for any claim A, including the present one, neither affirm A nor affirm
its contrary; and do not even affirm or deny that you are neither affirming nor
denying it. Of course, as anticipated, this cannot be taken as the formulation
of Academic Skepticism: first of all because, if it were, there would be a
proposition A (the formulation) that the Academic skeptic accepts; second,
because the formulation is clearly contradictory.
The three kinds of modal skepticism to be considered in this chapter all
propose that we suspend judgment with respect to (some of) our modal
opinions. They differ in the extent of the modal opinions that allegedly should
be doubted and in their motivations for doubting. Quinian modal skepticism is
a form of semantic doubting, because it advocates suspending judgment on
whether we can adequately analyze the meaning of certain modal sentences
(so-called de re modal sentences). Van Inwagen’s modal skepticism is of the
epistemic variety, because it advocates suspending judgment regarding
the truth of certain far-fetched scenarios (e.g. the scenario where I have a
mind but not a body). Radical skepticism, finally, embraces the full spectrum
of our modal talk: it advocates the suspension of judgment with respect to it
all.

9781472524263_txt_print.indd 50 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 51

§2.2. Quinian modal skepticism

If you were to ask an analytic philosopher to explain what skepticism is in
relation to modality, he would speak to you of the position long-defended by
Quine, one of the most influential philosophers in the United States in the
second half of the last century. Quine first put forward his modal skepticism
in a series of writings published in the late 1940s and early 1950s (1947,
1953a, 1953b), going on to refine the view for the subsequent fifty years
(cf. especially Quine (1974, 1976) and Barrett and Gibson (1990)). The main
lesson, drawn from these writings, is still a matter of debate (for some recent
contributions, see Fine (2005, Chapters 2 and 3), Divers (2007), and Keskinen
(2012)). Details aside, it is safe to say that Quine’s skepticism is far from being
as radical as the Academic skepticism of Arcesilaus, and is of a semantic
variety: it concerns the meaning of certain types of modal sentences. Quine’s
target was very specific and, in order to be appreciated, his modal skepticism
requires a bit of introductory discussion.
Quine’s polemical targets were certain sentences of quantified modal
logic (QML), as developed in particular by his Harvard colleague Clarence
Irving Lewis. QML is the cradle of possible-worlds semantics and it is for
this reason that Quine’s skepticism has been much debated: it struck at the
core of the most significant conceptual apparatus ever developed to analyze
modalities (for an overview of the modern origins of modal logic, see Ballarin
(2010)). At the syntactic level, QML uses the formal languages of predicate
(or propositional) logic, adding two operators, named after their shapes: one
operator for necessity—the so-called “box”—and the other for possibility—
the so-called “diamond.” At the semantic level, the standard version of QML
introduces one additional key element: a set of possible worlds, at which (or:
in which, depending on one’s views; see e.g. Williamson (2002)) the truth of a
sentence is determined. The set of possible worlds is crucial to clarifying the
truth conditions of modal sentences of increasing complexity. The analyses
of sentences involving only one of the two operators are, respectively, as
follows: “It is necessary that: ‘p’ is true when, and only when, ‘p’ is true at
all possible worlds”; “It is possible that: ‘p’ is true when, and only when,
‘p’ is true at at least one possible world.” Starting with these two analyses,
analyses of sentences involving multiple operators could then be provided.
So far, so good. As we know from the Introduction, many have quarreled
over the metaphysical status of possible worlds, but this was not until after
QML was established as a key conceptual tool for discussing modalities, in
the 1970s. Quine’s concerns are, rather, about the semantic details of the
theory. In particular, for Quine it is suitable to attempt to make an analysis
of certain de dicto modalities, but not of de re modalities. In order to outline

9781472524263_txt_print.indd 51 01/12/2015 09:44

 52 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

his position, we must therefore look into the classic distinction between de
dicto and de re in its connection with the topic at hand. Let us consider the
following sentence:

1 Foffo the cat eats fish and potatoes.

Next, let us distinguish three ways in which (1) can be involved in other
sentences. First, we sometimes find ourselves in a situation in which we
wish to say something with regards to (1). For instance, we might find Foffo’s
meal surprising, and so we might say:

2 It is surprising that: Foffo the cat eats fish and potatoes.

Those who assert (2) are not trying to add any detail to the situation described
in (1); on the contrary, they are assuming (1) and expressing a certain frame
of mind with regard to that situation. Similarly, in other cases, instead of a
frame of mind, we find ourselves expressing an attitude toward the situation,
or even a certain relationship between the situation being considered and
another. For example, we might say:

3 Norberto believes that: Foffo the cat eats fish and potatoes.

4 What Rosalba said calls into question that: Foffo the cat eats fish and
potatoes.

Let us keep in mind for a moment this first of three types of sentence,
in which a situation is related to a frame of mind, an attitude, or another
situation.
Second, at other times, we find ourselves wanting to add certain details to
the entire situation described in (1); for instance, we might say that:

5 Voraciously: Foffo the cat eats fish and potatoes.

Statement (5) does not express a frame of mind or attitude of the person
who makes it; nor does it express a relationship between situations; rather, it
modifies the situation described in (1) by adding a detail: it is now specified
that the meal took place in a certain way.
Third, at yet other times, we find ourselves wanting to add a detail that
does not concern the entire situation, but only some of its components. For
example, we might say:

6 Foffo the young cat eats fish and potatoes,

or:

7 Foffo the cat eats fresh fish and warm potatoes.

9781472524263_txt_print.indd 52 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 53

Quine (1953a) points out that these three ways in which a sentence can be
modified are also detectable when we consider modal terms—or, at least, when
we try to interpret their meaning. Let us consider the following three sentences:

8 It is possible that: Foffo the cat eats fish and potatoes.

9 Possibly: Foffo the cat eats fish and potatoes.

10 Foffo the cat is able to eat fish and potatoes.

Sentence (8) appears to be analogous to (2), (3), and (4) insofar as it expresses
a frame of mind of the speaker with regards to (1); (9) appears to be analogous
to (5), insofar as it adds a detail to the description of the entire situation in (1);
finally, (10) appears to be analogous to (6) and (7), insofar as it adds a detail to
Foffo, the cat, one of the components of the situation in (1).
Traditionally, sentences such as (2), (3), (4), and (8) are considered to be de
dicto: in them we say something about other sentences, without modifying
the situation that these sentences describe. Sentences such as (5), (6), (7),
(9), and (10) are considered to be de re: in them we modify the state of things
(res, in Latin); and “things” may refer to entire situations, as in (5) and (9), or
components of situations, as in (6), (7), and (10) (see Campbell (1964), Quine
(1953a), and (1953b)).
Incidentally, it is this last use of modal terms—illustrated in (10)—upon
which rest the doctrines of essence and capacity, or (as it has been called)
disposition, which we encountered in Chapter 1 and which will be discussed
again in Chapters 7 and 8. If we say that Giovanni knows how to drive a car,
or that he essentially belongs to the species Homo sapiens, we are attributing
a capacity and an essence, respectively, to Giovanni. This occurs through the
application of a modal term to one of the parts of the sentences: an operation
that is understood as a modification of the “thing” to which that part of the
sentence refers.
Getting back to our topic, the theory defended by Quine is that the only
legitimate use of modal terms is that illustrated in (8); sentences (9) and (10), by
contrast, use modalities in an equivocal manner. To frame Quine’s skepticism
within the present volume, recall the problem of possibility (PP), presented in
the Introduction, along with its three sub-problems (SPP, EPP, and MPP):

PP: What does it take for a certain situation to be possible?

SPP: What does it mean to say that a certain situation is possible?
EPP: How do we come to know that which is possible?
MPP: What sort of entity is a possible entity (of any given kind—an
individual, property, state of affairs, or …)?

9781472524263_txt_print.indd 53 01/12/2015 09:44

 54 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Quine’s view is that PP cannot be raised with respect to de re modalities.

Because sentences containing de re modalities—and, thus, the proposi-
tions these sentences express—contain equivocal terms, which cannot be
assigned a clear semantic function, we ought to refrain from employing such
sentences to express our (philosophical) opinions.
As for de dicto modalities, Quine’s solution lies in interpreting them as the
attitudes of the speakers toward the actual sentences. Such interpretation
is typically referred to as meta-linguistic, because it holds that modal claims
are assertions, in the language of the speaker, concerning claims of another
language—referred to as the object language. Thus, for instance, in (8) the
phrase “It is possible that” conveys the speaker’s attitude, while “Foffo the
cat eats fish and potatoes” belongs to the object language and is the target
of the speaker’s attitude. Consequently, for Quine a sentence containing
a de dicto modality will not express a possible situation, but rather the
perspective of the speaker toward a (non-modal) situation. As he summarized
it, “necessity resides in the way we talk about things, not in the things we talk
about” (1953a: 174). Incidentally, as we shall see, this motto is also largely
supported by expressivism.
The philosophical significance of Quine’s skepticism toward de re modal-
ities and his interpretation of de dicto modalities must not be underestimated.
They flew in the face of QML, the semantics fine-tuned by C. I. Lewis, which
would later be used to develop the possible-worlds semantics that is at the
core of this volume. Thus, if Quine is right, possible-worlds semantics has no
solid footing. But is he right?
To substantiate his theory, Quine appeals to considerations from both
metaphysics and the logical analysis of sentences. Let us first consider
the former. Put simply, if attributed to situations or components of them,
modalities can be very complex entities to explain. In particular, sentences
such as (9) and (10) force us to buy into Aristotelian essentialism, which
Quine regarded as highly suspicious. In the words of Quine, Aristotelian
essentialism holds that:

some of the attributes of a thing (quite independently of the language in

which the thing is referred to, if at all) may be essential to the thing, and
others accidental. E.g. a man, or talking animal, or featherless biped (for
they are in fact all the same things), is essentially rational and accidentally
two-legged and talkative, not merely qua man but qua itself.
—Quine (1953a: 175–6)

Aristotelian essentialism would, thus, force us to pinpoint, for each thing,

what is essential to it and what is accidental. For instance, we would have
to take a stance on whether Foffo is essentially furry or not, or whether his

9781472524263_txt_print.indd 54 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 55

genetic makeup is essential to him. But these are hard questions, to which
we should not presume that there are answers that are valid regardless of
how we approach them. To reiterate the point from a more familiar stance,
consider Napoleon. We all have heard of him and his deeds, but it is another
matter to be able to draw the line between what he did and what we believe
he could have done. Suppose we were to agree that he could have won at
Waterloo; how might we explain this attribution? And, how might we distin-
guish it from false attributions of possibility? For example, could Napoleon
have been a fisherman living in Ajaccio, France, in the eighteenth and
nineteenth centuries? Or, could he have been a tomato growing in Pisa in
May, 1977? Precisely in order to avoid troublesome questions of this nature,
which have pestered Aristotelian essentialism for centuries, suggesting the
need for an alternative, Quine refuses to regard assertions of the form of (9)
and (10) as meaningful.
The criticism arising from Aristotelian essentialism is connected to Quine’s
other reasons to distrust de re modalities, namely those concerning the
logical analysis of sentences. This is because essentialism is implied by
the logical analysis proposed by those who accept de re modalities:

My logical point about essentialism was that he who accepts quantification

into modal contexts as making good sense should not balk at essentialism
[…]. If you are going to take the one you must take the other. That was
not an argument against essentialism. But it happens further that I do not
myself make sense of essentialism, or of metaphysical necessity.
—Barrett and Gibson (1990: 244)

We have just seen Quine’s reasons to reject essentialism. His reference

to “quantification into modal contexts” is what we have to explain next. In
quantified modal logic, such quantification is of two varieties, corresponding
to sentences of the sort of either (9) or (10). First, we have quantification over
possible worlds (whatever these are taken to be), because necessity and
possibility are interpreted as expressing truth at (or “in”, depending on one’s
views), respectively, all worlds or some world. Thus, the sentence:

11 Necessarily, Foffo is furry

is analyzed as:

12 At all possible worlds, Foffo is furry.

While the sentence:

13 Possibly, Foffo is wet

9781472524263_txt_print.indd 55 01/12/2015 09:44

 56 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

is analyzed as:

14 At some world, Foffo is wet.

The second type of quantification concerns sentences where the modality

modifies a specific thing, as in (10). Thus, the sentence:

15 Foffo is essentially furry

is analyzed as predicating of Foffo the following:

16 There is an x such that, essentially, x is furry.

Now, in quantified modal logic “essentially” is standardly regarded as equiv-

alent to “necessarily” (see Chapter 8 for a challenge to this interpretation),
thus (16) is analyzed as:

17 There is an x such that, necessarily, x is furry.

Sentences such as (17) are the main target of Quine’s criticism. According
to him, because we have multiple ways of picturing any x (i.e. any thing),
the truth conditions of sentences like (17) will vary as our representation
of x varies. So, consider two descriptions of Foffo—the “x” in question in
(17)—such that both pick out Foffo, but one does so necessarily and the other
accidentally. If in (17) x is described as “Foffo the cat”, (17) turns out (at least
intuitively) to be true; but, if x is described as “Laura’s favorite pet”, (17) turns
out not to be true, because Laura’s favorite pet is furry (it is indeed Foffo),
but Laura happens to like cats only accidentally—she could have developed
a taste for lizards instead. The only way to avoid this problem is to buy into
essentialism, but we have seen that Quine does not want to do that.
Thus, according to Quine, by trying to take at face value de re modalities,
quantified modal logic is forced to adopt essentialism in order to remain able to
assign a fixed truth value to certain sentences. But essentialism is problematic.
Thus, we should not buy into quantified modal logic: “Necessity does not
properly apply to the fulfillment of conditions by objects […] apart from special
ways of specifying them” (Quine 1953b: 151). Now, Quine acknowledges that
certain assertions that are apparently de re, such as (9), can be reinterpreted as
de dicto assertions and can therefore be recognized as legitimate in some way.
Nevertheless, this type of de re assertion proves to be conceptually muddled;
how are we to explain, for instance, the meaning of the following sentence:
“Possibly: necessarily: possibly: Napoleon loses at Waterloo”?
To date, Quinian skepticism continues to play an important role in the philo-
sophical debate, as demonstrated by the recent works of Boghossian (2009),
Burgess (1997), and Neale (2000). Furthermore, it is crucial for understanding

9781472524263_txt_print.indd 56 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 57

the development of the debate on modality which we will consider in the

remainder of this book. After all, it was the desire to reject the skepticism
of their professor that drove Saul Kripke and David Lewis, Quine’s students,
each to formulate a semantics of de re modal expressions and a theory of
possible worlds, all of which have since structured the debate.
Needless to say, many philosophers did not buy Quine’s modal skepticism.
A particularly convincing line of reply in recent years has come from John
Divers (2007) (but see also Fine (2005: Chapters 2 and 3) for another reply
to Quinian skepticism). Divers considers the status of Quine’s criticism
of quantified modal logic after the works of David Lewis, that is, after the
“return of metaphysics” in analytic philosophy and, more specifically, in
the debates regarding alethic modality. To Divers, Quine equivocates between
semantic and metaphysical issues. Clearly, in the example discussed above,
how we come to describe “x” matters to the truth conditions of (17); and,
clearly, there is more than one way of describing x. Yet, this is not to say that
x, in itself, does not have necessary and accidental properties that belong
to it no matter how we describe it. The fact that we can describe the world
in many ways, and that what holds necessarily or possibly based on those
descriptions therefore varies, does not make the world itself inconstant and
dependent upon our descriptions. To put it in terms that are proper to this
volume, Divers criticizes Quine for conflating the SPP and the MPP. To agree
with Divers we have to endorse a good deal of metaphysics; but, as we will
see in later chapters (especially 4 and 5), there are multiple ways of rejecting
Quine’s modal skepticism, and gaining the conceptual benefits of quantified
modal logic, while refraining from committing to a wild metaphysics.

§2.3. Van Inwagen’s modal skepticism

A second variety of modal skepticism was put forward by Peter van Inwagen
(1998), in an article titled “Modal Epistemology”. While the article generated
considerably less discussion than Quine’s writings, it is, nonetheless,
instructive to discuss it because it defends a different sort of skepticism
regarding propositions involving modalities, namely epistemic skepticism.
Van Inwagen considers arguments that have been at the center of heated
disputes in the history of philosophy, such as the following three (18–20;
21–3; 24–6):

18 It is possible for there to be a being that has all perfections essentially.

19 Necessary existence is a perfection.

20 Therefore, There is a perfect being.

9781472524263_txt_print.indd 57 01/12/2015 09:44

 58 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

21 It is possible that I exist and nothing material exists.

22 Whatever is material is essentially material.

23 Therefore, I am not a material thing.

24 It is possible that there exists a vast amount of suffering for which

there is no explanation.

25 If there exists an omniscient, omnipotent, and morally perfect being,

there cannot also exist a vast amount of suffering for which there is
no explanation.

26 Therefore, It is impossible for there to be a necessarily existent being

that is essentially omniscient, omnipotent, and morally perfect.

Van Inwagen calls these possibility arguments because their first premise—
(18), (21), and (24)—claims that a certain situation is possible. A striking
feature of possibility arguments is that their logical structure seems sound;
and yet, many regard them as not valid. Moreover, it is extremely rare for
someone who is inclined to deny the validity of the argument to change
her mind at a later point. According to van Inwagen, arguments of this sort
suggest that our opinions regarding the truth of certain propositions involving
modalities rest on questionable grounds and we would do better to take a
skeptical stance towards them. This is because they describe far-fetched
scenarios about which we cannot have any opinion that we can suitably
justify. Here is a passage where van Inwagen expresses his view:

My own view is that we often do know modal propositions, ones that are
of use to us in everyday life and in science and even in philosophy, but do
not and cannot know (at least by the exercise of our own unaided powers)
modal propositions like the crucial modal premises of our three possibility
arguments. I have called this position “modal skepticism”. This name was
perhaps ill-chosen, since, as I have said, I think that we do know a lot of
modal propositions […] It should be remembered, however, that there
has been another sort of skeptic: someone who contends that the world
contains a great deal of institutionalized pretense to knowledge of remote
matters concerning which knowledge is in fact not possible. (Montaigne
was a skeptic in this sense, as were, perhaps, Sextus and Cicero.) It is in
this sense of the word that I am a modal “skeptic.”
—van Inwagen (1998: 69)

Van Inwagen’s skepticism challenges modal discourse at a crucial link in the

chain—that between possibility and imagination. Is imagination a reliable guide
to possibility? And, what does it take to imagine a possible scenario, really? The

9781472524263_txt_print.indd 58 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 59

discussion on this topic is vast and still ongoing. David Lewis (1986) famously
rejects that conceivability can be a reliable guide to possibility. A 1993 article
by Stephen Yablo—“Is Conceivability a Guide to Possibility?”—reopened the
debate. The article’s merit lies in the fact that it asks what the constraints
would be on our conception of conceivability in order for it to play a suitable
role as a guide to possibility. For instance, there are many scenarios that are too
complex for us to be able to conceive (e.g. those involving a transfinite number
of entities); or scenarios that are on the verge of being contradictory (e.g. the
scenario under which Goldbach’s conjecture is false); or scenarios that are too
iffy for us to take a stance (e.g. those indicated by van Inwagen). While we
cannot enter into the details of the debate, which tackles epistemic issues, it
is important to recognize that there are many propositions involving modalities
(including de re modal propositions) whose truth conditions we cannot establish,
even if we believe that the sentences expressing them are meaningful.

§2.4. Radical modal skepticism

We come now to the third type of skepticism, which is significantly more
naïve and—all things considered—radical than Quine’s and van Inwagen’s, in
that it encompasses the full spectrum of modal talk. For this reason, we will
name it radical modal skepticism. Although it would be doubtful to attribute
it to any particular author, nonetheless radical modal skepticism represents a
position that, from time to time, can act as an antagonist or as an ally in philo-
sophical disputes. More generally, to fully understand the theoretical force of
radical modal skepticism helps us comprehend the nature of each theory
of possibility. For this reason, it is very important to consider it in some detail.
Let us begin with a simple consideration. The majority of sentences that
concern possible situations deal with what can be termed mere possibilities:
situations about which there neither are, have been, nor will be any concrete
facts in our world. For instance, take the following sentence:

27 There could have been more stars than there are.

There are no facts or events in our world that indicate with certainty that
there could have been more stars than there are. How, then, are we to judge
the truth or falsehood of this affirmation? Why must we claim that, through
a theory, the truth conditions of (27) can be explained? Indeed, how are we
to make sense of the proposition expressed by (27)? Sentence (27) certainly
makes some sense in English, but so does the proposition:

28 Tomatoes fly only in February.

9781472524263_txt_print.indd 59 01/12/2015 09:44

 60 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

There are no facts in our world indicating whether tomatoes fly only in
February, so we are at a loss when trying to assign a truth value to (28).
Of course, these doubts can be extended to all other mere possibilities,
suggesting that only modal talk concerning actuality is meaningful. If we
concede this much, however, we probably can no longer plausibly maintain
that the PP (or, subordinately, the SPP, EPP, and MPP) has an answer. In
recent years, Huw Price (cf. 2004) has suggested that the problem posed by
mere possibilities be included in the list of what he calls placement problems.
The list includes all the cases that are hard to explain for those who are empir-
ically minded and believe that there is only the actual world. In the case of our
subject, how can such thinkers accommodate modal truths? Where are they
to locate the so-called truthmakers (i.e. the entities that would make-true) for
such truths? The radical modal skeptic interrupts the debate with an easy and
yet drastic solution: it would be better simply to exercise a radically skeptical
attitude towards modal talk.
It is very important to realize that radical modal skepticism does not cast
into doubt only abstruse cases. When she was stopped by the police, Giulia
was driving with a blood alcohol level that was higher than the limit allowed by
law; as a result, she will incur certain penal and administrative sanctions. Yet—
we might ask ourselves—what considerations justify such sanctions? One
crucial role pertains to certain merely possible events: driving with a blood
alcohol level that is higher than a given value makes the possibility of having an
accident (and, more importantly, an accident that involves people other than
Giulia) much more probable. Here a skeptical attitude, which sometimes stirs
within us, will set in: how can we incriminate Giulia on the basis of a possible
event that never occurred? How are we to know that Giulia could cause an
accident when she has a certain blood alcohol level, if she has never been
involved in one? Certainly, others have been involved in many accidents with
that same blood alcohol level, but not Giulia. And every person is different.
Little by little, we begin to see where the radical modal skeptic points
her finger: the difference between that which is and that which could have
been. This is effectively summarized at the beginning of chapter XV of Niccolò
Machiavelli’s The Prince (1513):

And many have imagined republics and principalities that have never been
seen or known to exist in truth; for it is so far from how one lives to how
one should live that he who lets go of what is done for what should be
done learns his ruin rather than his preservation.
—Chapter 15: 61

While Machiavelli’s objective was not to raise a general objection to the

theories of possibility, his position regarding the theory of effective government

9781472524263_txt_print.indd 60 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 61

adequately summarizes the radical skeptic’s crucial point. We would be crazy

to believe that we can justify not only our theories but also our life practices
on the basis of that which could have been, but which we know was not, is
not, and will not be. That which we cannot experience, we cannot judge or,
much less, rely upon.
The doubts raised by the radical modal skeptic involve the semantic,
epistemic, and metaphysical planes. With respect to the first, radical modal
skepticism can be taken to imply positions of increasing radicalness. In its
milder interpretation, radical modal skepticism argues of merely possible
scenarios that we cannot verify that their truth conditions are met, because
we can never witness them; nonetheless, we can tell what these truth
conditions are; thus, the propositions expressed by sentences about merely
possible situations are meaningful because we can understand what it would
take to verify them. Under a bolder interpretation, radical modal skepticism
claims that we cannot specify the truth conditions of sentences about merely
possible scenarios, since these sentences are really about entities that are
mysterious to us; thus, the question arises whether we even grasp the
meaning of the propositions expressed by these sentences.
As for the epistemic plane, it should be noted that radical modal skepticism
delves even deeper than skeptical attitudes adopted in other contexts, such
as the one that deals with inductive reasoning. The problem with inductive
reasoning is to make sense of an alleged link between past and actual events,
on one hand, and predictions of future events, on the other. If the future is yet
to come, then what will be seems much like what could be or—from after the
fact—what could have been, as opposed to what is or has been. (Incidentally,
this is the reason why temporal and modal semantics have been developed
alongside each other and are often still discussed jointly.) Nevertheless,
there is a crucial difference: predictions can be confirmed; of that which is
merely possible, though, we cannot have any sort of confirmation. For this
reason, we cannot attempt to justify a theory of possibility on the basis of
certain empirical correlations and their degree of confirmation, as some do
in the case of predictions. Thus, radical modal skepticism strikes deeper
than skepticism about induction and constitutes one of the chief epistemic
challenges to any theoretical project in modality. The conclusion is that “no
metaphysical account which renders it impossible to give a plausible episte-
mological theory is to be countenanced” (1985: 217), as Graeme Forbes
observes in his classic volume on modality.
As for the metaphysical plane, the suspension of judgment with respect
to propositions involving modalities, and in particular with respect to merely
possible propositions, is an ally of anti-realist approaches to modality. Our talk
of possibility, as it turns out, may regard no real entity. Indeed, many philoso-
phers who endorse or have endorsed some radical form of anti-realism, or

9781472524263_txt_print.indd 61 01/12/2015 09:44

 62 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

who look or have looked at metaphysics with little sympathy, will subscribe,
if not to the letter, at least to the spirit of radical modal skepticism. Mere
possibilities belong to the category of scenario that philosophers are better
off not discussing, never mind using as a basis for their theories.
There is a single way to respond to the radical modal skeptic: by asserting
that there are certain characteristics of our universe that can be understood
independently of experience, and that possibility is among these character-
istics. The only position that escapes this response is modal expressivism,
another radical but sui generis approach to modality. The rest of us, facing
the full force of the radical modal skeptic’s challenge, must appeal to that one
available response. This points toward the connection between that which is
possible and that which is actual, illustrating that the link cannot be found in
an empirical fact or the past, present, or future. The only response open to
theorists of possibility is to assert that the missing link (or links) is a semantic,
epistemic, or metaphysical fact whose validity must be understood and
accepted to some extent independently of experience.
First, we must clarify the expression “independently of experience”. Let us
consider the following sentence:

29 The sum of the angles of a triangle is 180 degrees.

When we say that (29) is independent of experience, we are qualifying the

way it can be learned. Without entering into superfluous detail, we will say
that a sentence is known independently of experience when there are at
least two ways (and typically many more than two) of learning the sentence
that are independent of one another. For example, Giovanni learned (29)
through a personal demonstration on May 10,1996 in Pisa; Rebecca learned
it from her elementary school teacher on June 12, 1987 in Philadelphia; Piero
just learned it by reading a book at home in Athens. In each of these cases,
the same proposition expressed by the sentence (29) is learned. And the
modes of learning are all independent of each other: in order for Giovanni’s
attempt to be valid, he does not have to be privy to what the person who
took him through the personal demonstration knows. By contrast, the
following statement requires a very specific observation in order to be
learned:

30 On April 27, 2008, at three o’clock in the afternoon, Dario Fo wears a

green shirt.

Certainly, Dario Fo himself can tell us how he is dressed, but our knowledge
depends on someone’s observation of the situation. There is no way to
become privy to (30) independently of the specific observation of the situation
depicted in (30) (with the exception of those who have divinatory powers).

9781472524263_txt_print.indd 62 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 63

Two qualifications are in order. First, independence of experience has

been regarded by some authors (e.g. Kitcher (1980, 2000)) as the hallmark
of so-called a priori knowledge. To the extent that this conjecture is correct,
then, we are suggesting that there is a core of modal epistemology that
necessarily relies on a priori knowledge because it cannot rest on the bare
observation of actual situations. Analogously, Humean skepticism toward
inductive reasoning pointed out that the bare observation of present events,
or recollection of past ones, cannot provide knowledge of a logically entailed
future event, indicating that there is a core of inductive reasoning that cannot
be justified a posteriori. Second, our suggestion is not that a broadly empiricist
account of modal knowledge cannot be provided (we will see some notable
examples in Chapter 3, on modalism, and Chapter 7, on the new actualism).
Rather, we are suggesting that a radically empiricist project, in which
knowledge of actual facts alone can justify knowledge of modal facts (and
thus of mere possibilities as well), seems destined to fail. Incidentally, the last
point may help explain why the most recent accounts of modal epistemology
(including Peacocke (1999), Chalmers (2002), Williamson (2007b), and Lowe
(2012)) all have a rationalist bent. However, as said, we shall not rule out too
quickly accounts that are more empiricist in spirit, such as that by Bueno and
Shalkowski (2015).
Let us recapitulate. In order to respond to the radical modal skeptic, we
must assert that certain aspects of possibility are learned a priori. Naturally,
the radical modal skeptic will argue that it is irrational to trust a priori
knowledge, and that it would be best to suspend our judgment regarding
sentences on possibility. Such is the key challenge that this radical form of
skepticism leaves us with. It is a challenge that must be taken seriously and
that affects each of the theories to be considered in the rest of the volume.
Most of those who have written about modality regard the radical skeptic’s
position as blind to an important characteristic of human knowers: some
aspects of reality are undoubtedly known a priori; and possibility, which is
omnipresent to us, is one of these. We cannot suspend our judgment of
possibility because we would at the same time be suspending our judgment
of reality. This is all fine: but what further response can we provide to the
radical modal skeptic’s challenge? Several answers will be considered in
the following chapters. Now, we shall proceed to see the position of those
who maintain that modal talk simply expresses emotions and who, for this
reason, do not feel the full force of the radical modal skeptic’s challenge.

9781472524263_txt_print.indd 63 01/12/2015 09:44

 64 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

§2.5. Modal expressivism

The first response to the skeptic we will consider is modal expressivism. The
expressivist position has been supported in various philosophical fields and
it is not particularly well-known as a modal theory. Perhaps the most fruitful
field of application for expressivism is meta-ethics, the branch of philosophy
that sets out to provide a definition of the concepts “right”, “wrong”, “good”,
and “bad”. The expressivist position is particularly significant in aesthetics, as
well. Among the classic texts in which expressivism of one kind or another
can be found, we might mention Spinoza (1677, esp. Appendix to Part I)
and Hume (1739–40). The authors who have most pursued this position in
recent times are Simon Blackburn (see especially (1984, 1993, 1998)), Alan
Gibbard (see especially (1990, 2003)), and Shroeder (2010, 2008a, 2008b). We
should also keep in mind the past formulations by Ayer (1936: Chapter 6) and
Wittgenstein (1953).
Expressivism is a peculiar theory of modality, in that it shares with radical
modal skepticism a strong anti-realist bent. The core claim of modal expres-
sivism is a denial that modal expressions bear cognitive content: they are,
rather, ways to express non-cognitive attitudes of speakers. We shall refer to
these attitudes as sentiments, interpreting the term broadly as referring to all
non-cognitive mental states.
Providing an adequate definition of non-cognitive can be complex; to
date, there is good reason to suspect that the distinction is not as clear
as it has appeared in the past (see, for example, Boghossian (1995)). In
any event, roughly speaking, non-cognitive states are those that lead us
to act, rather than to understand, a situation. If I believe that Giovanna
was wrong, this means (at least, roughly) that I possess a representation
of a situation in which she appears; yet it is not due to this that I will act
one way instead of another. If I wish to punish Giovanna, though, my wish
suggests a way in which I might act towards her—I might refuse to forgive
her. Beliefs are cognitive mental states, insofar as they essentially consist
of understanding a situation; desires, on the other hand, are non-cognitive
mental states, insofar as they are essentially not tied to a representation,
but to an action.
In the most recent and sophisticated versions of expressivism, sentiments
are understood against the background of social convention (cf. Gibbard
(1990, 2003)). Thus, the claim that a certain action is wrong is to be under-
stood as the speaker’s expression of conformity to a normative convention
regarding the action. The crucial feature of the account is that the expression
of conformity is a non-cognitive attitude: although beliefs or intentions can to
some extent influence it, ultimately such expression is rooted in a sentiment.

9781472524263_txt_print.indd 64 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 65

Analogously, modal expressivism takes sentences containing modalities to

express the speaker’s conformity to a certain conventional way of regarding
the non-modal content of the sentence. For example, the sentence:

31 It is possible that: Foffo will play with the carpet in the afternoon

will express the conformity, on the part of whoever endorses it, to all the
conventions typical of possibility (e.g. to leave open that in the future (31) will
be true, or to prepare for (31) to be the case, and so on).
By slightly modifying Quine’s motto, the expressivist theory could be
summarized as claiming that “necessity resides in the frame of mind with
which we relate to things, not in the things to which we relate”. A succinct
definition of the position, to be carried forward in the remainder of this
volume, is thus the following:

EXP: The sentence of natural language: “It is possible that: s” expresses

a certain (conventionally codified) sentiment, on behalf of the person who
asserts it, toward s.

The first and chief upshot of modal expressivism is that modal talk cannot
be charged with truth or falsity. Truth and falsity are based on situations, not
on sentiments; and the expressivist argues that sentences featuring modal
terms express sentiments, not beliefs. The expressivist, therefore, would
deny that the SPP, the EPP, and the MPP pose a problem. There is certainly
a captivating side to the expressivist proposal: explaining what it means to
say that a certain situation that is possible or necessary is an arduous task;
oftentimes, we are under the impression that there is no meaning in modal
expressions, and that they express an attitude of the speaker toward what he
is saying. Evidence of this is found in the number of times we use the term
“necessary”, relativized to the contexts of use: that which is necessary for the
mathematician; that which is necessary for a soccer player; or for a student;
or for Piero; or for Lulù, the dog; and so forth. The same can be said for the
use of the term “possible”. How are we to believe that there is one single
meaning to modal expressions? More importantly, how are we to believe that
modal expressions involve any meaning at all?
Two additional features of the modal expressivist theory deserve to be
highlighted. The first is that, unlike all other modal theories, the expressivist
has no need to reduce the complex variety of modal talk to the two standard
modalities of possibility and necessity. It is indeed plausible to conjecture
that there will be conventions associated with “It is possible that” distinct
from those associated with “It may be the case that” or “Perhaps” or “It is
likely that”. This is no small theoretical advantage, as keeping faith with the

9781472524263_txt_print.indd 65 01/12/2015 09:44

 66 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

pragmatic nuances of the variety of modal expressions is a daunting task for

basically all other modal theories on the market. Of course, such a feature
comes at the cost of giving up a cognitivist analysis of modality, and thus
giving up also the conceptual subtleties allowed by quantified modal logic.
The second feature is more delicate. Modal expressivism both can and
must offer an analysis of modal talk such that the non-cognitive attitude is
clearly demarcated from the cognitive component. For example, according to
expressivism, when we say “There could have been more stars than there
are” we are expressing a certain (conventionally codified) sentiment towards
the following non-modal sentence:

32 The number of stars is greater than the actual number.

In a way, therefore, modal expressivism is committed to interpreting all modal

talk in terms of de dicto modalities. This is because modal expressions are all
ways of speaking about a certain situation, and in particular they are all ways
of expressing certain sentiments regarding such situations (as when one says
“This ice cream is yummy!” or “Your behavior is disgusting!”). Contrary to
the Quinian skeptic, the expressivist would consider legitimate sentences
such as that illustrated in (10) (‘Foffo the cat is able to eat fish and potatoes’);
however, the expressivist would interpret these sentences as de re: that is
to say, as expressions of frames of mind addressed specifically to one of
the “things” (res) that belong to the situation expressed by the sentence.
It is nonetheless a challenge for modal expressivism to provide plausible
analyses for every sentence embedding some modality. For instance, it is not
straightforward to analyze claims embedding a predicate that itself involves
a modality, as in:

33 Clara is likeable.

It may prove equally difficult to analyze long and convoluted sentences

containing modal expressions, such as the famous example by Sobel cited in
Lewis (1973: 10):

34 If the USA threw its weapons into the sea tomorrow there would
be war; but if the USA and the other nuclear powers all threw their
weapons into the sea tomorrow there would be peace; but if they
did so without sufficient precautions against polluting the world’s
fisheries there would be war; but if after doing so, they immediately
offered generous reparations for the pollution there would be peace.

While the debate on expressivism in ethics is lively, very few are the recent
contributions to modal expressivism. Nonetheless, the ideas put forth with

9781472524263_txt_print.indd 66 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 67

regard to ethics may be used in coming years to provide an adequate level of

sophistication to modal expressivism as well.
Before considering some of the problems of modal expressivism, a variant
of this view deserves to be mentioned. It is the so-called modal normativism
put forward by Amie Thomasson (2010); on this topic, see also Sidelle (2009).
Modal normativism shares with modal expressivism the denial of the idea that
claims involving modality are descriptive claims that can be true or false. The
two views, however, depart on the analysis of the non-descriptive content:
while for the modal expressivist such content should be analyzed in terms
of specific sentiments, for the modal normativist “modal assertions serve a
normative (rather than descriptive) function of conveying constitutive rules
for using the terms” (Thomasson 2010: fn. 6). The chief philosophical task
for the modal normativist is therefore to specify the norms that are at play
in the constitutive rules of our modal expressions. These norms should not
be spelled out in terms of a logical framework such as the one provided by
QML. Rather, the goal is to clarify the rule which constitutes the foundation
for our use of modal expressions: that is, for each modal expression we should
specify when one is allowed to use it. For example, consider once again (8) (“It
is possible that: Foffo the cat eats fish and potatoes”): while a modal expres-
sivist reads (8) as conveying a certain non-cognitive attitude on the part of who
endorses it, a modal normativist analyses (8) in terms of the circumstances
under which a speaker is taken to reasonably claim, or an agent is taken to
reasonably believe, that Foffo the cat eats fish and potatoes.

§2.6. Two problems for modal expressivism

There are two principal criticisms of modal expressivism. The first concerns
sentiments; more specifically, we might ask ourselves: what sentiments are
we speaking of? The extant literature does not offer thorough answers on this
point. Still, we may attempt to clarify the matter for ourselves, starting from
the two paradigmatic cases of necessity and possibility.
Presumably, the sentiment associated with necessity is to be revealed in
the absolute steadfastness with which one accepts the sentence in question.
If we believe that, necessarily, the sum of two plus two is four, this means
that we would be willing to make a lofty bet on the truth of the sentence,
or that we are ready to fervently defend the truth of the sentence from those
who claim that the sum is five. The sentiment associated with possibility,
on the other hand, leads to a break: if we believe a certain situation to be
possible, we would also be willing to entertain the thought that things might
be or have been otherwise.

9781472524263_txt_print.indd 67 01/12/2015 09:44

 68 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Nevertheless, bringing sentiments into play when dealing with possibility

arouses certain perplexities. In fact, a given sentence can be the object of
varying sentiments, depending on the occasion. Let us consider, for example,
how sentiments vary in relation to how likely we believe it is that the
sentence will be confirmed. We will assign the sentences that are considered
necessary the same likelihood (or the value 1.0, which indicates certainty)
and, therefore, we can reasonably expect a similar frame of mind; but when
we consider the sentences that are held to be possible, we will have values
that span the entire range of probability, from greater than 0 to 1.0 (because,
after all, that which is necessary is also possible). Let us take the following
sentences:

35 It is possible that you are reading this sentence.

36 It is possible that a recipe for happiness will be discovered by next April.

37 It is possible that next year I will win New Year’s raffle.

38 It is possible that the sum of two plus two is four.

It appears difficult to claim that there is one, univocal sentiment that is

associated with these sentences. And, besides, this would not be sufficient:
in order for the modal expressivist theory to be correct, the sentiment must
be specific to all and only sentences concerning possibility.
Earlier, we had mentioned that modal expressivist theory might be more
appealing than the cognitivist approach to modality: the objection might be
made against the latter that a single meaning of expressions such as “possible”
and “necessary” does not exist. However, this criticism raises a similar problem
for the modal expressivist: it does not seem plausible to assert that there is one
sentiment expressed by all and only sentences that refer to a possibility.
In reality, the modal expressivist is worse off than the cognitivist about
modality. Here we might introduce the second criticism, which can initially be
expressed circumstantially and later generalized. Relationships of implication
between meanings can be constructed with relative ease. For example, if
certain dietary elements are necessary for all soccer players, then they will
also be necessary for Giovanni, who is a soccer player; or, if it is logically
necessary that a whole be greater than each of its parts, then it will also be
physically, biologically, geometrically necessary. Yet, on the other hand, it has
been verified that it is difficult to find ties between sentiments. We might
infer that a person who likes apples also likes apple juice; if she likes to travel,
then she would most probably like to receive an airplane ticket to Marrakech
as a gift; but, unfortunately, this is often not the case.
The question of inferences raises a more general problem for the
expressivist theory—one that concerns all forms of expressivism (see

9781472524263_txt_print.indd 68 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 69

the discussion of this point in Hale (1993)). There are certain inferences
among modal sentences that are considered valid by the vast majority of
human beings. Let us take, for example, the following argument:

39 a is a triangle.

40 It is necessary that: if x is a triangle, then x has three sides.

41 Therefore, It is necessary that: a has three sides.

The validity of this argument appears rather simple to demonstrate.

Nevertheless, an expressivist encounters serious difficulties when explaining
why the argument is valid. In fact, in order to do so, a theory of implication
between modal sentiments is required: yet, as we have seen, there is good
reason to believe that this sort of theory cannot be provided.
Moreover: a proponent of the cognitivist approach would point out that she
can also explain the validity of inferences that are far more complex than the
one just seen, on the grounds of the meanings attributed to the two principal
modal concepts: “necessary” and “possible”. Let us consider, for example,
the following argument:

42 a is a triangle.

43 It is necessary that: a is identical to b.

44 It is necessary that: if x is a triangle, then x has three sides.

45 For every triangle x, it is possible that: x is isosceles.

46 Therefore, It is necessary that: b has three sides, and it is possible

that: b is isosceles.

A cognitivist can explain the validity of this argument by resorting to logic;

but not an expressivist. The expressivist would retort at this point that the
concept of validity cannot be applied to arguments containing modal expres-
sions. Or, more specifically: we cannot apply the concept of validity implicit
in the criticism raised, which concerns the relationship between a series of
propositions that guarantees the truth of the conclusion, upon the acceptance
of its premises. This is because expressivism denies that there is truth or
falsity when it comes to sentences involving modalities.
However, this does not mean that the expressivist is unable to explain
the acceptance of an argument that contains modal terms. At least two
explanations can be conceived. According to the first, the acceptance of an
argument is explained through the use of a certain sentiment with regard to
the conclusion of the argument itself, or, the sentences (41) and (46) in the
two previous examples. The second explanation introduces an expressivist

9781472524263_txt_print.indd 69 01/12/2015 09:44

 70 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

definition of validity as a sentiment we express when we consider the

relationship between a series of sentences; that is to say: we can have senti-
ments not only with regard to single sentences, but also certain relationships
between sentences.
Yet the temptation remains to believe that necessity and possibility involve
something more than mere sentiments. Sentiments are not the only thing—
we might believe—that make us quit smoking in the face of health effects or
that lead us to punish or praise a person based on what he could have done.
There’s something more: an actual argument founded on the understanding
of certain modal concepts. And it is not an argument that we have made up
through the introduction of a constitutive rule, as the modal normativist would
have it. It is just these modal concepts that we often find combined with the
representation of a situation, and that express what we call possibility and
necessity. The following chapters will be devoted to the theories that aspire
to explain these presumed concepts.

Study Questions
• What is the main target of Quinian modal skepticism?
• What is the distinction between de dicto and de re sentences?
• Why are de re modal sentences problematic for Quine?
• What is quantified modal logic?
• What is the main target of van Inwagen’s modal skepticism?
• What is the main challenge posed by radical modal skepticism?
• What is the definition of modal expressivism?
• Name three distinguishing features of modal expressivism.
• What are the main problems for modal expressivism?
• What is modal normativism?

FURTHER READING
The distinction between de dicto and de re attitudes, propositions, and sentences
plays a key role in several philosophical fields. For the applications to modality, see
especially: Quine (1953a, 1953b); Campbell (1964); Lewis (1979, 1986: Chapter 4);
Fine (1974, 1994); and Burgess (1997).
Among the views examined in this chapter, Quinian skepticism is the most
debated. For an overview of the position from the point of view of the debate on
modality, see Haack (1978: Chapter 10), Forbes (1985: 216–17), and Melia (2003:
Chapter 3). See Hylton (2007) for an overview that places Quine’s skepticism

9781472524263_txt_print.indd 70 01/12/2015 09:44

 MODAL SKEPTICISM AND MODAL EXPRESSIVISM 71

within the context of his other work. Quine’s position is most famously outlined
in his work of 1953a but, for a more comprehensive picture, it is important to look
at other writings as well, including those of 1947, 1953b, 1974, 1976, and also
Quine’s contributions to Barrett and Gibson (1990), in particular his replies to Ruth
Barcan Marcus. Critical appraisals of Quine’s position abound and, to a certain
extent, all the theories discussed in Chapters 3–7 of this volume can be regarded
as rejoinders to Quine. For a reconstruction of Quine’s view, see Divers (2007) and
Keskinen (2010, 2012). Among the most recent articles directly discussing Quine’s
position, see Burgess (1997), Neale (2000), and Ray (2000). Classic articles that
gave shape to the debate include: Smullyan (1948); Føllesdal (1961); Parsons (1967,
1969); the essays collected in Linsky (1974); van Fraassen (1977); and the essays
collected in Marcus (1993: especially 14 and 15). For a reconstruction of the main
moments that characterize the modern origins of modal logic, see Ballarin
(2010).
Van Inwagen’s modal skepticism is outlined in his work of 1998. For a recent
defense of the position, see Hawke (2011). Van Inwagen’s position should be
read against the background of the broader debate on the relationship between
conceivability and possibility; some important readings in this regard are: Yablo
(1993, 2001), Sosa (2000), Geirsson (2005, 2014), Kung (2010, 2011), Roca-Royes
(2011), and Ichikawa and Jarvis (2012).
Radical modal skepticism cannot be attributed strictu sensu to any
particular author. For a formulation of the placement problem, also with respect
to modality, see Price (2004). On the a priori, see Casullo (2003) and Russell
(2014). For a relatively recent and influential collection of papers, see Boghossian
and Peacocke (2000). For some recent work on the epistemology of modality, see
Peacocke (1999), Chalmers (2002), Williamson (2007b), Lowe (2012), and Bueno
and Shalkowski (2015).
For a short introduction to modal expressivism, see Melia (2003: Chapter 1).
For an overview of the expressivist position with a special focus on its applications
in ethics, see Shroeder (2010) and van Roojen (2014). For a defense of expressivism
in ethics, which might be employed also to elucidate modal expressivism, see
Shroeder (2008a, 2008b) and, especially, Gibbard (1990, 2003). Modal expressivism
was notoriously endorsed by Blackburn in a series of writings, including his works
of 1971, 1984, 1993, and 1998. Among the important precursors of the position
are: Spinoza (1677: Appendix to Part I); Hume (1739–40); Ayer (1936: Chapter 6);
Wittgenstein (1953) and the related discussion in Wright (1980: Chapters 19–23).
For some recent articles, see especially Hale (1993), Sidelle (2009), and Thomasson
(2010).

9781472524263_txt_print.indd 71 01/12/2015 09:44

9781472524263_txt_print.indd 72 01/12/2015 09:44
 3
Modalism

A general characterization of modalism is first provided. Modalism accepts that

modal statements can be true or false and, moreover, that the formal language
of QML is well-suited to express them; however, modalism refuses to provide a
semantics that further analyzes the modal operators of necessity and possibility,
such as a possible-worlds semantics (§3.1.). Next, two problems for modalism are
discussed: one related to the expression, within the modalist framework, of certain
types of sentences involving modalities; the other concerning the relationship
between the formal language adopted by modalists and natural language (§3.2.).
Three more recent variants of modalism are then considered, concerned with—
respectively—the semantics, metaphysics, and epistemology of the theory. The
semantic variant was proposed by Charles Chihara: it brings together the modalist
credo that modal concepts cannot be analyzed and the use of possible worlds to
provide truth conditions for modal sentences of QML (§3.3.). The metaphysical
variant is inspired by Jonathan Lowe’s proposal that modal knowledge rests on
our knowledge of essences (§3.4.) The epistemic variant, which aims to make
empiricism possible for modalists, was put forward by Bueno and Shalkowski
(§3.5.). The chapter ends with a look beyond modalism, by means of which
we introduce the notion of possible-worlds semantics for QML as providing a
conceptual analysis of modal expressions (§3.6.).

*****

9781472524263_txt_print.indd 73 01/12/2015 09:44

 74 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

§3.1. At the roots of modalism

B ased on what has been discussed in the previous chapter, from here on
out we will analyze the positions of those who: (i) accept the challenge
posed by the radical modal skeptic to show that modal expressions (including
de re expressions) express something; and (ii) dismiss the hypothesis that
modal expressions express sentiments.
The next area of dispute concerns the analysis of the modal concepts
presumably expressed by modal terms. In other words: is it possible to
undertake an analysis of modal concepts? The so-called modalists claim that
it is not; proponents of possible worlds (modal realists, ersatzists, fiction-
alists, and agnostics) and dispositionalists claim that it is. In this chapter, we
will focus on the modalists. Lastly, we will be introduced to the concept of
possible world, which will appear frequently in the remainder of the book.
Modalism takes its name from the acceptance of modality as a brute fact.
That is to say: the modal concepts “possible” and “necessary” express the
possibility or necessity of a situation. Period. While it may seem trivial, this
position merits further consideration. In order to explain it, we will turn to a
well-known metaphor in contemporary metaphysics: the universal catalogue.
Let us suppose that we are to compile a catalogue of the universe, in which
we will include everything—every last thing—that exists. We will include
Napoleon and the Battle of Waterloo; the Colosseum; Tasmania; rhythm and
blues; cider; and so on. At some point, we will have to consider that which
is possible: should we include it in the catalog, and, if so, how? The modal
skeptic would advocate doubting that modal scenarios can be included; the
expressionist would include the sentiments expressed by modal terms;
and the modalist? She would include … modal facts: the possible victory
of Napoleon at Waterloo; the possible end of the world on December 31,
1999; the necessary fact that the sum of two plus two is four; and so forth.
Therefore, the modalist’s solution to the PP is the following:

M: The sentence in natural language: “It is possible that: s” expresses

the existence of a modal fact of our world: the fact that it is possible
that s.

Let us consider, for example, the following sentence:

1 Today, in Pisa, Foffo the cat could have eaten milk and cookies.

According to the modalist, (1) expresses a fact of our world, which is compa-
rable to the fact that there are three apples on this table: the fact that today,
in Pisa, Foffo the cat could have eaten milk and cookies.

9781472524263_txt_print.indd 74 01/12/2015 09:44

 MODALISM 75

Despite the simplicity of the theory, modalism is the result of serious

reflection on modal concepts and the analysis that can be made of them
through a formal theory. Precursors to modalism can be found among
Ancient, Medieval, and Modern philosophers, including Aristotle. But it is
with the advent of quantified modal logic (QML) that modalism as such
can be identified as a theory on the market. While modalism need not be
expressed by means of the language of QML, it typically is. In other words,
modalists typically do not set out to apply the principle M directly to the
terms and modal concepts of ordinary language, but to the well-formed
formulas of QML, in particular those formulas where the symbols expressing
necessity (i.e. the box: “£”) and possibility (i.e. the diamond: “¯”) appear.
As noted in the Introduction, the modalist’s choice to work with a formal
language is a common one among the theories discussed in this book.
Not only modalism, but modal realism, ersatzism, fictionalism, and agnos-
ticism prefer to treat questions of the metaphysics of possibility through a
language in which the meaning of expressions has been established in an
unequivocal manner. One of the motivating reasons behind this choice is the
ease with which a notion of truth and one of validity, applicable also to modal
sentences, can be established through an artificial and unequivocal language.
As seen earlier, the expressionist must accept that judgments regarding what
is possible or necessary are all, to some extent, equivocal. It is not easy to
understand what sentiments can be used to define possibility or necessity;
nor is it a simple task to identify what relationships exist between senti-
ments. The use of an artificial and unequivocal language eliminates these
problems.
The core idea for a modalist is to endorse the syntax of QML, but reject
the semantic account appealing to possible worlds. A sentence such as:

2 It is necessary that: Foffo is a cat

is not true because, at every world, “Foffo is a cat” is true; rather, it is a brute
fact of the actual world that (2) is true. Analogously, the sentence:

3 It is possible that: Foffo has milk and cookies

is true not because, in at least one world, Foffo has milk and cookies;
rather, it is a brute fact of the actual world that (3) is true. If this is the case,
however, modalists must be able to tell a story—in particular, a metaphysical
story—regarding actual primitive modalities. That is, although resorting to
the language of QML eliminates ambiguities and helps solve one of the
three core questions presented in the Introduction, the two other questions
remain—even more urgently—open. Recall the questions:

9781472524263_txt_print.indd 75 01/12/2015 09:44

 76 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

SPP: What does it mean to say that a certain situation is possible?

EPP: How do we come to know that which is possible?
MPP: What sort of entity is a possible entity (of any given kind—a
possible individual, property, state of affairs, or …)?

For the modalist, the answer to the SPP is trivial. To say, for example, that it
is necessary that Foffo is a cat is to say that it is necessary that Foffo is a cat;
this is because the necessity of such a scenario is a brute fact. Yet, can the
modalist provide suitable answers to the EPP and MPP? This is the central
question to be addressed in this chapter.
An early modalist proposal is found in the postscript to Prior and Fine
(1977). The first to defend modalism in a systematic manner was Graeme
Forbes (1985, 1989), and his work remains fundamental for those interested
in this theory. We will come back to Fine’s view in Chapters 7 and 8. For
the time being, in order to pinpoint two problems for modalism, we shall
assume the perspective of Forbes (§3.2.). Afterwards, we shall consider
three important alternative versions of modalism, focusing respectively on
its semantic, metaphysical, and epistemic aspects. In §3.3., we will consider
the variant proposed by Charles Chihara (1998), according to which the
semantics of modalism can make use of possible worlds, thus enjoying their
conceptual benefits, while denying that necessity and possibility can be
analyzed in terms of possible worlds. In §3.4., we will consider the variant
suggested by Jonathan Lowe (2012), which introduces the metaphysical
category of essences in order to provide a suitable interpretation for the
operators of necessity and possibility. Finally, in §3.5., we will discuss the
proposal of Octavio Bueno and Scott Shalkowski (2015, 2013, 2009), whose
ambition is to put forward a modalist view with an empiricist outlook. The
chapter will end (§3.6.) with a look ahead to theories that, unlike modalism,
endorse a semantics for QML based on possible worlds.

§3.2. Two problems for modalism

In Forbes’s classic rendition, modalism rests on three choices: (i) enriching
the language of first-order logic with the two operators for necessity and
possibility; (ii) maintaining—unlike modal skepticism and expressivism—that
sentences containing these operators do have a truth value; but also (iii)
holding that it is a matter of brute fact whether a sentence is necessarily
or possibly true or false. Declining to provide an analysis of modal expres-
sions (in QML), the modalist avoids the metaphysical and epistemic troubles
of possible worlds, which we will discuss in the next three chapters; on

9781472524263_txt_print.indd 76 01/12/2015 09:44

 MODALISM 77

the other hand, modalism runs into objections. We will examine two main
ones. The first regards the expressive power of the theory, the second the
relationship between, on one hand, the two modal primitives countenanced
by the language of the theory and, on the other, the issues in natural language
that the theory is supposed to address. Let us consider them in order.
In numerous works (1992, 2003, 2005), Joseph Melia has pointed out that
the modalist finds it difficult to translate certain sentences into the artificial
language of the theory. Consider, for example, the following sentence:

4 There could have been more stars than there are.

Sentence (4) makes a comparison between the stars of this world and
those of a possible world. More generally, (4) makes a comparison between
two situations: actuality and a possible way that actuality might have been.
Because the modalist does not allow for there to be more than one world, it
is not clear how she can express a comparison between actuality and what
might actually have been. The problem had already been discussed in Forbes
(1989). There, Forbes sets out to enrich the language of the theory with
terms that carry out the same function as a possible world. That is, he intro-
duces a term An for every world Wn, where the term’s function is to relativize
the interpretation of a sentence (or one of its parts) to a certain context.
Thus, Forbes can relativize a part of (4)—the part that concerns the number of
stars in our world—to the actual “actuality”, and the other part—the part that
concerns the possible number of stars—to another “actuality”. The strategy
is effective from a formal point of view; but authors such as Melia argue that
the meaning of the “actuality” expressions is obscure unless it is interpreted
in the spirit of possible-worlds semantics, jeopardizing the modalism of the
theory.
Other paradigmatically challenging sentences are those referring to
multiple possibilities. Let us consider the following:

5 There are three possible ways to secure a contract.

How is a modalist to explain the connection between the three possible ways
of securing a contract? The translation would say that there are (only) three
actualities, in each of which there is a way to secure a contract. Or, that the
three possibilities are each relativized to a different actuality. But, it’s hard to
see what the actualities mean without appeal to a metaphysics of possible
worlds. Once again, the strategy of relativizing different actualities produces
a result that is satisfying from a formal perspective, but troubling from a
metaphysical and epistemic point of view.
Fortunately, Melia offers another way of interpreting (4) and (5). This form
of interpretation differs from Forbes’s in proposing a dissociation between

9781472524263_txt_print.indd 77 01/12/2015 09:44

 78 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

the artificial language and the world it speaks of. In order to explain the
meaning of (4) and (5), it is not necessary to provide a paraphrase of them in
the artificial language of the modalist theory or to make an explicit reference
to the relationships between the various situations (the three possibilities in
(5) and the actual and possible number of stars in (4)). A simple list of the
situations will suffice; for comparison, a metaphysical explanation (expressed
in ordinary, not artificial, language) is provided. Consequently, (4) would be
interpreted in the following manner (and it should be imagined that the inter-
pretation can be expressed in formal language as well):

6.1 In the actual world, there are x number of stars;

6.2 it is possible that: the number of stars is greater than x.

Statements (6.1) and (6.2) are both easily expressible in formal language:
the former does not contain any modal terms; the latter is a simple modal
sentence. The relationship between the two sentences must be explained in a
supplementary manner by means of natural language (e.g. by pointing out that it
is the very same entities and kinds of entities, viz. stars, that are under scrutiny).
And (5) would be interpreted in the following way (again, it should be
imagined that the interpretation can be expressed in formal language):

7.1 It is possible that: the contract is secured in manner A;

7.2 it is possible that: the contract is secured in manner B;
7.3 it is possible that: the contract is secured in manner C.

Statements (7.1), (7.2), and (7.3) are all expressible in formal language: they are
nothing but three simple sentences, each expressing a possible fact.
The solution proposed by Melia closely follows a methodology supported
primarily by David Armstrong, according to which an artificial language can
be useful, but cannot completely substitute for natural language when facing
a philosophical problem. (See, for example, Armstrong (1997) and (1978).)
In the contemporary debate on possible worlds, this is certainly one of the
most profound points of discussion. Further ahead, in Chapters 4–6, we will
analyze the four theories that oppose Armstrong’s methodology; not until the
final chapter, when we discuss dispositionalism, will we see a theory that is
inspired by it.
We shall now consider the second criticism aimed at modalism, which
concerns the relationship between the formal language in which the theory
as interpreted by Forbes is molded and the three problems of possibility—the
SPP, EPP, and MPP.
There are entire books that discuss the metaphysics of possibility through
the almost exclusive use of an artificial, unequivocal language. Yet the

9781472524263_txt_print.indd 78 01/12/2015 09:44

 MODALISM 79

problem lies in being able to connect the expressions of such a language with
those of ordinary language. For example, as illustrated earlier, the language
used by Forbes makes use of the following two symbols: “£” and “¯”.These
symbols are meant to capture to some extent the meaning of the concepts
of necessity and possibility as expressed in natural language. If they were
not, Forbes would not claim to have resolved the problem of possibility, but
simply that he had provided us with a new language. Still, many disagree with
Forbes. Some—whose positions we will soon examine—believe that the two
symbols must be analyzed further. Others believe, more simply, that Forbes’s
work does not actually address the three central questions introduced at the
beginning of this book. Let us take a brief moment to see why.
Fundamentally, there are two reasons to doubt that the symbols “£” and
“¯”, together with a series of rules concerning their correct use, capture
what we mean when we talk about possibility and necessity. The first is that
Forbes, like the other modalists, does not provide a precise model for trans-
lating from the artificial language of the theory into ordinary language. Forbes
claims that modal facts are among the facts of our world. But, seeing as
his language only offers us two symbols to express these facts, it becomes
difficult to interpret certain sentences in natural language, such as:

8 Perhaps Ubaldo could have and should have eaten less ice cream
last night.

The point is not that we lack a model for translating (8) into the language of
Forbes’s theory (although this sort of objection can be made for sentences
with distinct characteristics, as we will soon discover), but that Forbes’s
theory does not jibe with the way we speak of possibility. According to
Forbes’s theory we must interpret (8) as the expression of three facts: that
Ubaldo ate ice cream; that he could have eaten a smaller portion than the
one he did; that he should have eaten a smaller portion than the one he
did. Yet many will read (8) as the expression of a single fact: that Ubaldo
ate ice cream—a fact with which a certain possibility and certain obligations
are associated. Perhaps, then, modalism does not yield a metaphysics of
modality; it is unable to provide an interpretation of our modal expressions
that would explain what makes sentences containing such expressions either
true or false. And, in this sense, modalism does not solve the MPP.
The second reason for doubting the explanatory power of the symbols
“£” and “¯” lies in what is customarily referred to as their brute character. It
could be argued that the modalist does not give us a theory of possibility since
he does not believe he must explain the meaning of modal expressions. As
we shall see in the following chapters, a theory of modality is not required to
provide a definition of the principal modal concepts (possibility and necessity):

9781472524263_txt_print.indd 79 01/12/2015 09:44

 80 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

that is to say, it is not necessary to find concepts that can completely replace
modal concepts, as is the case, for instance, when we define the water
molecule in terms of two atoms of hydrogen and one of oxygen arranged
in a certain way. Nevertheless, a theory must at least explain the principal
modal concepts by means of some kind of philosophical analysis: that is, it
must demonstrate how the relevant concepts are tied to other concepts that
are familiar to us. The modalist denies the need for any definition or expla-
nation; the only requirement is to identify the modal concepts through certain
symbols of a formal language that avoid the equivocality of ordinary language
expressions. To many, this is an utterly unsatisfactory solution that trivializes
both the EPP and the PP.

§3.3. Modalism with possible worlds

Over the past two decades, some important variants of modalism have been
put forward. They can be grouped together insofar as they share the core
idea that facts about modality are brute; however, they also differ in important
respects. The first variant we will consider was proposed by Charles Chihara;
it focuses on the semantic aspect of modality.
Chihara sets out his position primarily in the volume The Worlds of
Possibility (1998), which is divided in two parts. In the first, Chihara criticizes
four main views. Three of them appeal to possible-worlds semantics; they
are the views of David Lewis, Alvin Plantinga, and Gideon Rosen, which we
will discuss respectively in Chapters 4, 5, and 6. The fourth view is Forbes’s
modalism.
Lewis’s and Plantinga’s theories can be regarded as realist theories of
possible worlds, in that they read possible-worlds semantics as talking
about real entities, namely, possible worlds. On the other hand, Rosen’s and
Forbes’s theories are anti-realist. That said, they are anti-realist for different
reasons, which is important to specify in order to understand why Chihara’s
view falls in the modalist camp. Rosen is a fictionalist about possible worlds:
he regards the conceptual analysis of modality in terms of possible worlds as
well-suited to the ends of modal theorizing, but he rejects the metaphysical
commitment to worlds by arguing that possible-worlds semantics is a useful
fiction; possible worlds are no more real than the world of Shakespeare’s
Othello. Just as the play is a story that deeply moves and enthuses us
without being about any real thing in the world, so possible worlds have the
theoretical virtue of explaining modal scenarios without being about any real
thing (more on fictionalism in Chapter 6). Forbes is an anti-realist in quite a
different sense. Forbes takes modal expressions in the language of QML to

9781472524263_txt_print.indd 80 01/12/2015 09:44

 MODALISM 81

be about brute modal facts. While Rosen endorses the analysis of modality
by means of possible-worlds semantics, but attempts to stay away from
its dubious metaphysical implications, Forbes provides no analysis of modality
at all.
Chihara’s position, which he sets out in the second part of his volume,
resembles Forbes’s more than Rosen’s. Chihara, in fact, denies that it is
possible to provide a conceptual analysis of modal expressions in QML,
much less of those in natural language. For this reason, Chihara’s can be
regarded as a modalist position. On the other hand, Chihara maintains that
modalism is compatible with the acceptance of possible-worlds semantics.
This is because a schema can be provided that takes us from possible-worlds
semantics to a semantic theory that does away with possible worlds. Chihara
goes into great detail to spell out and defend his view. The core idea exploits
a standard trait of semantics for non-modal sentences: it is relatively standard
to evaluate the truth of a non-modal sentence according to an interpretation
of its terms, which specifies what the terms are about; thus, for any given
sentence, there will be a set of admissible interpretations for its terms. For
instance, consider the sentence:

9 Foffo is happy.

There will be an interpretation according to which (9) is about Foffo, the cat;
another interpretation according to which (9) is about a certain Foffo living
in Paris; another still, according to which (9) is about a certain Foffo living in
Lima; and so on. In the case in point, that is, each interpretation will provide
a different value for “Foffo”, so that the truth of (9) may vary depending on the
interpretation. Chihara’s suggestion is to regard possible-worlds semantics as
analogous to truth-under-an-interpretation, where the actual world plays the
role “Foffo” plays in (9), being reinterpreted as (or: substituted by) one world
or another, depending on the scenario. Thus, in the sentence:

10 It is possible that: Foffo is happy

the actual world is substituted by a possible world such that, at that world,
“Foffo is happy” is true. The semantic machinery of possible worlds turns
out to be an exercise of substituting other worlds for the actual world, and
evaluating sentences in accordance with such substitution. No conceptual
reduction is really achieved with the addition of possible worlds: we have
simply made the semantic operation of substitution more familiar and easy
to follow.
Chihara’s proposal is the outcome of a thorough study of the semantics
of modality; and it is ingenious. It definitely puts on the map a non-realist
account of modality that seems to satisfy Forbes’s desiderata without falling

9781472524263_txt_print.indd 81 01/12/2015 09:44

 82 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

prey to the conceptual shortcomings of Forbes’s view. On the other hand, as

pointed out by Sider (2001), Chihara overlooks important metaphysical and
epistemic desiderata for a theory of modality. To be fair, Chihara does not
seem to be concerned with those desiderata and it is still an open question
whether his proposal could be adequately supplemented to meet them. Any
modal theory should at least provide an indication of how sentences within
the formal language of the theory are linked to sentences of natural language;
it should also explain how the meaning of those sentences is grasped by
agents; and it should, finally, explain what those sentences are about. Thus,
until Chihara’s view answers these demands, we cannot deem it a satis-
factory account of modality.

§3.4. Modalism and essentialism

Metaphysical and epistemic aspects of modalism are at the core of our next
two modalist proposals. Modalism has been regarded by many as metaphysi-
cally suspicious because it introduces brute modal facts, without any story
of what grounds such facts. In a 2012 article by Jonathan Lowe—“What Is
the Source of Our Knowledge of Modal Truths?”—we find the inspiration for
providing adequate metaphysical foundations to the modalist project. Lowe’s
proposal is to find such a foundation in the modal entities known as essences,
which—according to him—we have independent reasons for admitting into
our metaphysics. Once the existence of essences is recognized, we gain
access to a suitable epistemic account of our modal knowledge.
Before considering Lowe’s proposal in more detail, it is important to dispel
a potential source of taxonomic confusion. Lowe’s project—as he himself
claims (e.g. 2012: 934)—was inspired by Kit Fine’s work on essences. Fine’s
view, though, is more generally considered as part of the so-called New
Actualism, which will be discussed in Chapter 7. This is because—unlike the
modalists—Fine argues that the two operators of necessity and possibility
fall short of providing adequate conceptual richness to a theory of modality;
instead, he advocates the adoption of a variety of “necessity operators”,
among them an operator for essential dependence. Thus, Lowe’s project can
be regarded as belonging to modalism only to the extent that essences are
invoked to provide the metaphysical foundation for a QML language equipped
with the two classic box and diamond operators (expressing, respectively, the
concepts of necessity and possibility). For this reason, we will rather say that
Lowe inspired—rather than endorsed—a certain metaphysical foundation for
modalism.
Lowe sums up his view as follows:

9781472524263_txt_print.indd 82 01/12/2015 09:44

 MODALISM 83

Put simply, the theory is this. Metaphysical modalities are grounded in

essence. That is, all truths about what is metaphysically necessary or
possible are either straightforwardly essential truths or else obtain in
virtue of the essences of things. An essence is what is expressed by a
real definition. And it is part of our essence as rational, thinking beings
that we can at least sometimes understand a real definition—which is just
a special kind of proposition—and thereby grasp the essences of at least
some things. Hence, we can know at least sometimes that something is
metaphysically necessary or possible: we can have some knowledge of
metaphysical modality. (2012: 947)

There are (at least) three major theoretical claims summed up in this passage.
The first concerns the metaphysical foundation of modality, which is found
in essences. The second is a doctrine of essences as real definitions. The
third is the view that our own essence as thinking beings entails that we are
able to understand at least some real essences. Let us look at each of those
claims, in order.
Modalists such as Forbes and Chihara saw modalism as a form of anti-
realist theory of modality, that is, a theory with no additional metaphysical
commitments other than the commitment to the existence of the two modal
concepts utilized in QML. The modalist variant inspired by Lowe’s proposal,
by contrast, sets out to find truthmakers for modal sentences of QML. Of
course, such truthmakers will have to belong to the actual world, as there is
no other world. Lowe, moreover, suggests that the truthmakers are concrete,
spatio-temporal entities. They are essences. Because it sees our modal
claims as being about concrete, spatio-temporal entities, the modalist variant
under consideration is a hardcore realist view.
Secondly, Lowe’s account of essences departs from classic accounts (e.g.
Kripke (1980)) based on possible worlds. The latter explicates essences in
terms of essential properties. Essential properties are those that belong to an
individual in every possible world at which the individual exists. The essence
of an individual is therefore defined as the totality of its essential properties.
Kit Fine criticizes this view because it is not sufficiently fine-grained. For
instance, arguably all of us have the essential property of living in a world
such that “2 + 2 = 4” is true at that world; or, Foffo, the cat, has the essential
property of belonging to the set that has Foffo as its sole member; but, intui-
tively, such properties should not be part of (respectively) our essences or
Foffo’s essence.
To avoid the pitfalls of the classic view, Lowe (taking his cue from Fine and
classic texts such as John Locke’s Essay on Human Understanding) proposes
to regard essences in terms of real definitions of a thing:

9781472524263_txt_print.indd 83 01/12/2015 09:44

 84 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

A real definition of an entity, E, is to be understood as a proposition which

tells us, in the most perspicuous fashion, what E is—or, more broadly,
since we do not want to restrict ourselves solely to the essences of
actually existing things, what E is or would be. (2012: 935)

The key feature of the view that essences are real definitions is that it makes
essences identical with the entities they define. Foffo’s essence is Foffo and
not—as the account of essences in terms of possible worlds has it—a cluster
of Foffo’s properties. An essence, in other words, is nothing over and above
the individual that it defines. This is a quite nice result for the modalist variant
under consideration: Lowe’s suggestion, it turns out, commits the modalist
simply to the existence of individuals. Thus, for instance, once we have Foffo,
we also have Foffo’s essence and, thus, we will have sufficient metaphysical
foundation for all the modal claims regarding Foffo.
At this point, our reader may suspect that Lowe has done some
metaphysical juggling. How is it possible that the actual Foffo provides the
foundation for all modal claims about Foffo, without buying into the existence
of any further entity? How can it be that Foffo, who ate fish and potatoes
today—that very Foffo—is the foundation for the contention that Foffo could
have had milk and cookies instead? Foffo did not have milk and cookies, so
there is nothing in the actual Foffo that entitles us to conclude that he could
have had a different meal.
Lowe’s rejoinder to this objection rests on the third of his claims in the
foregoing passage: that our own essence as rational, thinking beings entails
that we are able to understand at least some real essences. This is the core of
Lowe’s proposal in modal epistemology. It is a rationalist contention, for Lowe
holds that we need not experience—say—Foffo having cookies and milk in
order to understand that such a situation is possible for Foffo; as long as we
grasp Foffo’s essence, we also grasp that possibility.

Given that all metaphysical modality is grounded in essence, we can have

knowledge of metaphysical modality, provided we can have knowledge
of essence. Can we? Most assuredly we can. […] Knowing an entity’s
essence is simply knowing what that entity is. And at least in the case of
some entities, we must be able to know what they are, because otherwise
it would be hard to see how we could know anything at all about them.
How, for example, could I know that a certain ellipse had a certain eccen-
tricity, if I did not know what an ellipse is? (2012: 944)

Thus, according to Lowe, in order to know that Foffo could have milk and
cookies, we do not even need to see, or somehow be acquainted with, Foffo.
All it takes to have that knowledge is to know that Foffo is a cat.

9781472524263_txt_print.indd 84 01/12/2015 09:44

 MODALISM 85

Lowe’s proposal rests on two controversial bases: a rationalist perspective

on knowledge and a view of real definition that hardly seems extendable
to all entities (e.g. what is the real definition of bucatini all’amatriciana?).
Nonetheless, Lowe’s work inspires a plausible story for a modalist and realist
theory of modality, a story that deserves being placed on the map of the
theories of modality.

§3.5. Modalism and empiricism

Finally, we come to the third modalist proposal. Its distinctive trait is to be
geared towards philosophers of an empiricist bent. Empiricism is the view
that sensory experience is a necessary requirement for knowledge; nothing
can be known, that is, unless it is first experienced through the senses. The
problem of possibility is especially daunting for an empiricist because, as we
noted in Chapter 2 when discussing radical modal skepticism, by definition
we cannot experience mere possibilities (those scenarios that could have
happened, but did not). What is a modalist who is also an empiricist going
to say about our knowledge of mere possibilities? And what grounds such
possibilities in actuality?
Bueno and Shalkowski’s main contention is that modal knowledge rests on
judgments of similarity and the observation of constant regularities:

In both the well-defined and the less well-defined cases, judgments about
a particular case are made against a background of similar cases, constant
regularities, and the like so that we can have warranted judgments about
particular outcomes based on track records for similar cases. (2015: 688)

We are now far from Lowe’s rationalist proposal, where knowledge of Foffo’s
possible meal was derived from our a priori understanding of Foffo’s real
essence. Bueno and Shalkowski, rather, suggest that we know that Foffo
could have had milk and cookies because we have seen other cats like
Foffo eating such meals, or because cats regularly have milk and cookies.
Conceivability plays no role here: what matters is actual observation.
In order to argue for their main claim, Bueno and Shalkowski set out to
provide a “geneaology of modal knowledge”—that is, to explain how modal
knowledge is developed from particular observations. They begin with how
simple possibility claims (such as the one regarding Foffo) are known starting
from actuality claims. Then they consider dispositional claims (e.g. “Foffo’s
cage is fragile”), necessity claims, and probability claims. Their project is quite
ambitious because, arguably, it is unclear how all our modal knowledge can

9781472524263_txt_print.indd 85 01/12/2015 09:44

 86 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

be grounded in the observation of actual events. We shall come back to this

issue in Chapter 7, when discussing dispositionalism. Despite the challenges
faced by their proposal, Bueno and Shalkowski have secured an epistemic
foundation for the anti-realist modalist, which had been missing since the
works of Forbes.

§3.6. Possible worlds

At this point, before going on to discuss the theories that endorse a
conceptual analysis of modal concepts, it will be helpful to introduce the
notion of so-called possible worlds. As we have seen, the typical modalist
expresses the core of the theory in the formal language of QML, equipped
with two symbols—“£” and “¯”—that express the concepts of necessity
and possibility, respectively, of ordinary language. Nevertheless, the modalist
proposal has proven to many to be unsatisfactory, especially because it is
unable to explain the two principal modal concepts.
The modalist choice to use the formal language of QML while analyzing
claims containing the box and diamond as being about the actual world is far
from standard. In fact, since at least the 1950s, a semantics for QML based
on possible worlds has been predominant. In Chapters 4, 5, and 6, we will
see a number of different metaphysical positions advanced to back up such
a semantics. Each of these positions is problematic and, today, more than
fifty years later, views that do away with worlds are coming back (and will be
discussed in Chapter 7). But if, despite the difficulties, the metaphysics of
possible worlds has been pursued for at least half a century, this is because
it rests on a formidably nuanced and conceptually powerful semantics. Thus,
before moving forward with our discussion, some remarks on possible-
worlds semantics are in order.
While it was in its embryonic stages in the works of Wittgenstein (1921)
and Carnap (1946, 1947), possible-worlds semantics underwent a significant
evolution starting at the end of the 1950s with the works of the young Kripke
(1959, 1963) and Hintikka (1961, 1963; see also the reconstruction of Copeland
(2002)). Possible-worlds semantics is a theory of the meaning of modal
expressions, expressed in a formal language, in which certain terms are
referred to as “possible worlds”. Intuitively, a possible world is a thorough
and consistent scenario, a way in which our world could have existed, though
there could be worlds that are so different that they have not a thing to
do with our own. Simply put, the theory sets out to evaluate the truth and
falsity of every situation expressed by a modal sentence of natural language
according to the following schema (where s stands for any sentence):

9781472524263_txt_print.indd 86 01/12/2015 09:44

 MODALISM 87

(P) The sentence in natural language: “It is possible that: s” is true if and
only if the sentence in the language of the possible-worlds theory:
“There is a possible world w in which: s” is true;
(N) The sentence in natural language: “It is necessary that: s” is true if
and only if the sentence in the language of the possible-worlds theory:
“In every possible world w: s” is true.

In other words, possible worlds represent all of the possible scenarios that
we can imagine. If we say (in English) that a situation is possible, in the
language of the possible-worlds theorist we are saying that there is a world
(a scenario) in which that situation is realized; if we say (in English) that a
situation is necessary, in the language of the possible-worlds theorist we are
saying that, in all worlds (all scenarios) that we can imagine, that situation is
realized. For possible-worlds semantics, thus, every English sentence talks
about a world—either the actual world, or some of the possible worlds. For a
systematic formal presentation, see Girle (2000, 2003). Also useful is Garson
(2006).
The apparatus of possible-worlds semantics is standardly completed with
the addition of a relation among worlds, the so-called relation of accessibility,
defined for each world. (The semantic theory introduced by David Lewis and
discussed in Chapter 4, counterpart theory, is a notable exception to this rule:
in counterpart theory the key relation is not accessibility among worlds but
qualitative similarity among worlds.) Each world, that is, is assigned a set of
worlds, which are accessible from it: intuitively, this is the set of worlds that
define the full spectrum of possibilities that hold at that world. There are no
constraints on how the relation of accessibility is defined. The set of acces-
sible worlds may be empty—those will be the worlds at which nothing is
possible. Or it may contain only the world for which the accessibility relation
is defined—those will be the worlds at which that which actually happens,
necessarily happens. Or, it may contain all the possible worlds whatsoever:
those are the worlds at which that which is possible coincides with the
maximum extent of possibility. Typically, each world will have access to a set
of worlds in which some characteristic is fixed, such as the set of worlds
that contain the same laws of nature, and that is smaller than the total set of
possible worlds.
By means of this simple semantic apparatus, possible-worlds semantics
manages to express conceptual nuances that extend far beyond our linguistic
intuitions. For instance, consider the following sentence:

11 It is possible that: it is necessary that: it is possible that: Foffo the cat

speaks English.

9781472524263_txt_print.indd 87 01/12/2015 09:44

 88 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

While (11) is conceptually tortuous in English, possible-worlds semantics

delivers a straightforward machinery to check its truth conditions:

12 There is at least one world Wx accessible from the actual world and
such that, for any world Wy accessible from Wx, there is at least a world
Wz accessible from Wx such that, at Wz: Foffo the cat speaks English.

Of course, (12) is still conceptually tortuous, but checking the truth value of (11)
now becomes a matter of doing the proper calculation in terms of what’s true
at certain possible worlds. Possible-worlds semantics turns modal discourse,
as a variety, into a discourse about possible worlds and accessibility relations
among these worlds, a discourse that can easily be implemented and checked
by a machine. Indeed, possible-worlds semantics has undergone immense
growth in its fifty years of life; it has been revised, enriched and fine-tuned
in countless ways to give rigorous and straightforward expression to an
incredible spectrum of topics, including: mathematical claims (e.g. theorems
and proofs); claims in the natural sciences (e.g. laws and principles); medical
claims (e.g. epidemics and predictions); the behavior of agents in game
theory and economics; ethical principles; political principles; engineering
systems; and more. Today, we can claim that the conceptual machinery of
possible-worlds semantics enabled an analysis of the various modal expres-
sions, and of their conceptual ties, that is much more profound than the
analyses provided by any other society or civilization up until this point.
Nevertheless, the problem encountered with modalism presented itself
again with possible-worlds semantics: how are the terms that express
modality in QML to be interpreted in natural language? In other words: what
are these possible worlds the theorist of possible-worlds semantics speaks
of? Joseph Melia nicely sums up the problem:

Many philosophers have been impressed by possible-worlds semantics.

Possible-worlds semantics has shed light upon a wide range of modal
notions, has helped us understand counterfactuals and answered questions
about the validity and soundness of modal logics. But many philosophers
have been depressed by possible worlds themselves. Impressive though
its scope and explanatory power be, Lewis’s theory that merely possible
worlds are concrete objects like our own simply defies belief, and the
ersatzer’s theories are not particularly palatable to those with a taste for
desert landscapes. But without accepting some form of possible worlds
into our ontology how can we regard possible-worlds semantics as anything
more than a formal game? And if the semantics is nothing more than a
formal game then how could it shed any light upon our modal notions?
—Melia 2000: 333

9781472524263_txt_print.indd 88 01/12/2015 09:44

 MODALISM 89

The criticism of the semantic notion of possible worlds deserves particular

attention. Certainly QML, as developed by theorists of possible-worlds
semantics, has tremendous conceptual potential. Yet the key philosophical
problems that interest us here concern our everyday behavior; furthermore,
those problems are rooted in the language and concepts that we use daily,
not in the language of particular specialists. When we ask what the meaning
of modal expressions is, or what a possible entity is, or how we are to know
that a certain situation is possible, we ask in English, in Finnish, or in Tagalog;
and the answer must be in that language, not in a formal language.
Some (not the majority) consider this criticism to be misplaced. After all,
the main topics of medical, technological, or engineering research concern
our everyday behavior and are rooted in ordinary language, and yet we
have gained great benefits from the development of specialized languages
(including formal languages) for the consideration of these themes. Who, as
a layman, can sit through a specialized seminar on cardiology or electronic
engineering and truly understand what is being discussed? Why, then, should
philosophers be required to express their theories in ordinary language? Or,
why should they care about the metaphysical or epistemological implications
of their theories, so long as those theories serve to provide rigorous analyses
of certain ways of speaking?
It could be replied that, in order to carry out their work, a doctor or an
engineer might find it sufficient to use ordinary language when writing
a medical prescription or when explaining how to use a certain piece of
machinery. The rest of their work is conducted outside language—in actual
treatment of patients or dealing with machinery. The case of philosophical
research seems to be different: is the philosopher’s pill not made of words?
How can we expect a philosophical theory to resolve a problem if it does not
offer an answer in the language in which the problem was raised? As Sider
(2001: 91) points out, among the goals of a theory of modality there is also
providing “an account of quantification over possibilities in everyday language,
for example ‘there are five ways to win this chess match’.”
At this point, the dispute shifts towards the purpose of philosophy: perhaps
philosophy does not attempt to resolve the problems of ordinary humans, but
only of those who are willing to venture into the formal languages of philoso-
phers and who are able to perceive the solutions offered through them. The
question extends beyond the themes of this book but, having read up until
this point, perhaps the matter is worth pondering. A good start is provided in
Hendricks and Symons (2005).
And yet, implicitly or explicitly, the majority of analytic philosophers who
have concerned themselves with possibility have accepted the challenge: to
explain in their own language the meaning of what is known in formal language
as possible worlds. We shall take this up in the following three chapters by

9781472524263_txt_print.indd 89 01/12/2015 09:44

 90 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

examining four (types of) theories of possible worlds. It should be kept in mind
that the term is nothing more than a figurehead: just as the term “God” has
been identified with all that there is, with the supernatural, with history, and
with a specific person—to offer a few examples—the term “possible world”
has been identified with a potentially infinite number of entities.
What, then, is a possible world?

Study Questions
• What is the relationship between modalism and QML?
• What is the core thesis of modalism?
• What are the two main objections to Forbes’s modalism?
• What is the distinctive trait of Chihara’s modalism?
• What are the three distinctive theses of Lowe’s modal account?
• What is an essence, according to accounts based on possible worlds?
• What is an essence, according to Lowe?
• How do we know the essences of individuals, according to Lowe?
• What is the distinctive trait of Bueno and Shalkowski’s modalism?
• What is possible-worlds semantics?

FURTHER READING
For an overview of modalism that stresses the semantic issues of the theory, see
Melia (2003: Chapter 4). For an early version of the view, see the postscript to Prior
and Fine (1977). A classic exposition is found in Forbes (1985, 1989). Chihara’s variant
is amply developed in Chihara (1998). For Lowe’s essentialist proposal, see Lowe
(2012); for an introduction to theories of essence, see Robertson and Atkins (2013). For
Bueno and Shalkowski’s position, see Shalkowski (1994) and Bueno and Shalkowski
(2009, 2013, 2015). Among the writings that set out to discuss specific versions of
modalism, see: Sidelle (1989), Melia (1992), Forbes (1992), and Melia (2005).
For an overview of the issues surrounding the relationship between logic and
metaphysics, a good starting point is Hendricks and Symons (2005); see also
Zalta (2010).
For a historical overview of possible-worlds semantics, see Ballarin (2010)
and Copeland (2002). For an introductory treatment of the topic, see Menzel
(2015). A systematic presentation of the semantics is offered in a large number of
volumes and handbooks, among them, Girle (2003). Among the classic texts that
gave shape to this branch of logic and formal semantics are C. I. Lewis (1918),
Wittgenstein (1921), C. I. Lewis and Langford (1932), Carnap (1946, 1947), Prior
(1957), Kripke (1959, 1963), and Hintikka (1961, 1963).

9781472524263_txt_print.indd 90 01/12/2015 09:44

 4
Modal realism

M odal realism is the most bizarre contemporary take on modal metaphysics

and one of the most debated. It is associated with the name of David Lewis
and it was a pivotal contribution to the field, reviving interest in metaphysics
among analytic philosophers. First, a general characterization of modal realism
is provided (§4.1.). Then, its core metaphysical (§4.2.), semantic (§4.3.), and
epistemic (§4.4.) tenets are presented. Five theoretical virtues of modal realism
are then considered (§4.5.). Finally, we consider a major objection to the modal
realist project, according to which the theory’s effort to reduce modalities to the
existence of individuals (in worlds) fails on two counts (§4.6.).

*****

S ometimes referred to as “extreme realism” or “concretism” (see Divers

(2002), Melia (2003: Chapter 5) and van Inwagen (1986)), modal realism
is the theory of possibility conspicuously associated with the name of David
Lewis. Sketched in the article “Anselm and Actuality” (1970), its first formu-
lation is in Counterfactuals, published in 1973. But Lewis’s views were fully
developed in On the Plurality of Worlds. The latter volume, published in 1986,
is one of the most influential studies in analytic metaphysics of the past fifty
years; it is focused on the metaphysics of modality and has thus contributed
decisively to the legitimacy of this field of metaphysics. The importance of
Lewis’s work on modality can hardly be overestimated; since 1986 the debate
on the metaphysics of modality and cognate issues has been one of the
liveliest in analytic metaphysics and has paved the way for other theories of

9781472524263_txt_print.indd 91 01/12/2015 09:44

 92 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

possibility, such as modal fictionalism, which we will discuss in Chapter 6.

And yet, the view defended by Lewis is patently bizarre, so much so that he
himself claimed (1986) that modal realism produced an “incredulous stare” in
the person who first heard about it.
To date, only a handful of philosophers have openly endorsed modal
realism, a recent and notable case being Yagisawa (2010). Our challenge in
this chapter is to understand why a philosopher as brilliant as David Lewis
defended such a bizarre view, and whether it was worth it. We begin with
an outline of modal realism (§4.1.), followed by an exposition of its core
metaphysical (§4.2.), semantic (§4.3.), and epistemic (§4.4.) tenets. Five
theoretical virtues of modal realism are then considered in §4.5. Finally, in
§4.6., we consider a major objection to the modal realist project, according to
which the theory’s effort to reduce modalities to the existence of individuals
(in worlds) fails on two counts.

§4.1. The realist program

Modal realism literally follows the advice, given by Robert Musil and cited
at the outset of the Introduction, “to attach no more importance to what is
than to what is not”. To Lewis, possible scenarios and the possible entities
they involve exist in exactly the same sense that we would say that you and
I, the Amazonian forest, and the Eiffel tower exist. According to Lewis, there
are infinite worlds, as real or concrete as we would be willing to say that our
world is real or concrete. Such a simple and bold conjecture invites a wide
variety of philosophical objections, yet solves in an elegant manner the three
sub-problems of possibility.
Lewis’s theory is a brilliant exercise in philosophical analysis. The argument
for modal realism is presented as resting on the fruitfulness of the hypothesis
that there is an infinity of worlds as real as our own. In an often-cited passage
from the initial pages of On the Plurality of Worlds, Lewis defends his philo-
sophical stance based on its theoretical consequences:

Why believe in a plurality of worlds?—Because the hypothesis is

serviceable, and that is a reason to think that it is true. The familiar analysis
of necessity as truth at all possible worlds was only the beginning. In the
last two decades, philosophers have offered a great many more analyses
that make reference to possible worlds, or to possible individuals that
inhabit possible worlds. I find that record most impressive. I think it is
clear that talk of possibilia has clarified questions in many parts of the
philosophy of logic, of mind, of language, and of science—not to mention

9781472524263_txt_print.indd 92 01/12/2015 09:44

 MODAL REALISM 93

metaphysics itself. Even those who officially scoff often cannot resist the
temptation to help themselves abashedly to this useful way of speaking.
(1986: 3)

We shall return to the theoretical advantages of modal realism shortly.

To begin to explain Lewis’s choice, we must point out how modal realism
elegantly solves the three sub-problems of possibility, which—as we may
recall—are:

SPP: What does it mean to say that a certain situation is possible?

EPP: How do we come to know that which is possible?
MPP: What sort of entity is a possible entity (of any given kind—a
possible individual, property, state of affairs, or …)?

To answer the semantic problem of possibility (SPP), Lewis developed

counterpart theory, to be discussed in §4.3. Counterpart theory is a semantics
for modal sentences, resting on the existence of an infinity of worlds, such
that there is no individual inhabiting more than one world. Introduced in
1968—so, before modal realism—counterpart theory is the main alternative
to the standard possible-worlds semantics (developed by Hintikka, Kripke, et
al.), in which the very same individual may inhabit infinite worlds. As for the
metaphysical problem of possibility (MPP), Lewis’s solution is straightforward:
a possible entity is as real as any entity of the actual world. Finally, Lewis’s
answer to the epistemic problem of possibility (EPP) rests on philosophical
analysis: we know of the existence of other worlds and their inhabitants
because conjecturing that they exist delivers the best philosophical analysis of
modality, that is, a simple and elegant analysis with great explanatory power.
To position modal realism on the map of metaphysical theories of modality,
it is useful to compare it to other views. Let us consider the following
example:

1 There could have been more stars than there are.

The modal realist, contrary to the radical modal skeptic, believes that it is
possible to grasp the meaning of (1); and, contrary to the modal expressivist,
believes that (1) is not the expression of a sentiment of the person who uses
it, but is about a genuinely possible scenario. According to the modal realist
(and the modalist, for that matter), the analysis of (1) must begin by bringing
to light the modal operator: that is, there must be a certain relationship
between (1) and:

2 It is possible that there could be stars that actually do not exist.

9781472524263_txt_print.indd 93 01/12/2015 09:44

 94 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

For the modal realist (who here parts ways with the modalist), the truth of
(2) is best couched in terms of possible worlds: modalities can be concep-
tually analyzed in terms of worlds. Finally, the modal realist project is the
boldest of those that endorse possible worlds: this is because, according
to Lewis, modal realism not only conceptually analyzes modalities in terms
of non-modal entities (i.e. worlds), but also metaphysically reduces the first
to the second; as we shall see in due course (§4.5.), however, whether the
reduction succeeds is a point of contention.
According to the modal realist, there are many worlds, all of which are as
concrete as our world; the individuals inhabiting those worlds (people, living
entities, material objects, and any other entity whatsoever) exist in the same
way we say the inhabitants of our world exist. Of these worlds, there is one
in which there are more stars than there are in ours. Thus, (2) can be refor-
mulated in the following manner:

3 There is a world in which there are more stars than there are in
our world.

Notice that (3) is not about “possible worlds”, but “worlds”: this is because
all worlds exist on a par, as we shall see. Therefore, for the modal realist,
believing (1) simply amounts to believing (3). Generally speaking, the central
tenet of modal realism—through which the PP is resolved—is the following:

MR: The sentence in natural language: “It is possible that: s” expresses

the existence of a world in which: s.

The extravagance of the proposal is shocking. In order to explain the attri-

bution of possibilities to our world, we are left with infinite additional concrete
situations. To assess the plausibility of this view, let us now take a closer look
at the main conceptual focal points of modal realism. We shall begin with
the metaphysical underpinnings of modal realism, since they are the most
controversial and the most central to our analysis. Afterward, we will discuss
the semantic and epistemic sides of modal realism, in that order.

§4.2. Modal realism: Metaphysics

The main thesis defended in On the Plurality of Worlds—that there are infinite
worlds, each of which exists like our own—regards modal metaphysics in
particular. And yet, in order to argue for such a thesis, Lewis unfolds a much
richer metaphysical picture, where modal issues are entrenched with bits of
mereology (the theory of parts and wholes), as well as with a metaphysics

9781472524263_txt_print.indd 94 01/12/2015 09:44

 MODAL REALISM 95

of individuals, properties, time, and much more. Introducing modal realism,

therefore, forces us to touch upon a number of additional metaphysical
issues that have been at the center of contemporary debates in analytic
metaphysics. Here we have made a selection of the key issues and, even so,
we will not be able to discuss all of them in detail. We begin with the modal
realist’s conception of worlds.

§4.2.1. Worlds
Call the universe we inhabit, whose boundaries coincide with the boundaries
of its spatio-temporal manifold, the actual world. According to the modal realist,
infinite worlds exist. Some depart from the actual world in only minute details,
whereas some are so different that we cannot even imagine them. For instance,
there is (at least) one world differing from the actual just because, in the previous
sentence, instead of “minute details” the reader finds “small ways”: there is a
world inhabited only by philosophers; there is a world where libraries exist and
nothing else (no writers, no readers, no trees); and there is a world where only a
vast amount of Hydrox-cookie ice cream exists. Play as much as you like the game
of picturing worlds: as long as there is no contradiction in your picture, that world
exists. But, again, do not form the impression that only the worlds we can imagine
exist: there may be many more (we shall come back to this in the next section).
Each of the worlds of the modal realist possesses three key features: it is a
maximal individual, it is isolated from all other worlds, and it is not metaphysi-
cally privileged, or unprivileged, with respect to any other world. One may
add a fourth feature, already mentioned, which is that worlds are “concrete”.
Indeed, Lewis claims:

So, by and large […] it seems that indeed I should say that worlds as I take
them to be are concrete; and so are many of their parts, but perhaps not
all. But it also seems that to say that is to say something very ambiguous
indeed. (1986: 86)

So, yes, worlds exist in the same sense that we would say the actual world
exists and, at least in this sense, they are concrete; if people in the actual
world are made of flesh and blood, then other worlds contain people made
of flesh and blood too. But, because of the ambiguities in the meaning
of “concrete”, it is best not to place too much emphasis on the claim that
worlds are concrete. We will see that the key claim is, rather, that no world is
metaphysically privileged, or unprivileged, with respect to the others.
The first feature, then, is that each world is a maximal individual. It is an
individual because, no matter how gerrymandered a world may be, it is just a

9781472524263_txt_print.indd 95 01/12/2015 09:44

 96 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

totality composed of its inhabitants; like a salad is composed of its ingredients,

no matter how disparate they are, so a world is composed of its inhabitants.
In fact, for Lewis a world is nothing over and above its inhabitants, in virtue
of the (controversial) mereological principle of Composition as Identity, which
he accepts, and according to which, once you have the inhabitants of a world,
you have the world, and vice versa.
A world is a maximal individual because, in first approximation, it fully
occupies a reticulate of external relations. At least for our world and the
worlds like ours, such reticulate is a spatio-temporal relation—a specific
relation for each world. The definition of maximality resorts to a maximal
external relation, rather than a spatio-temporal system, because the former is
a wider concept than the latter: indeed, the former, but not the latter, can be
used to characterize worlds with a space-time manifold wildly different from
the one we attribute to the actual world.
In more detail, the first feature delivers the following condition of
maximality, which an individual must satisfy in order to count as a world:

I–MAX: w is a maximal individual if and only if all the parts of w bear a

certain external and maximal relation to all the other parts of w.

This condition depends on the following two types of relations:

R–EXT: A relation R between two individuals x1 and x2 is external if and

only if there are (at least) a part p1 of x1 and (at least) a part p2 of x2
such that: p1 and p2 do not overlap and R holds between p1 and p2.
R–MAX: A relation M is maximal with respect to a collection x1,…, xn if
and only if, for any xi among the x1,…, xn, xi bears the relation M to all
the x1,…, xn and there is no xj such that: xj is not among the x1,…, xn
and bears M to one of the x1,…, xn.

The second feature of each of Lewis’s worlds is that it is isolated: no

individual inhabits more than one world. Put together with maximality, the
resulting picture has it that the inhabitants of each world are unique to it
and there is one external relationship through which they are all connected.
Isolatedness may seem an innocuous feature, but it is distinctive of Lewis’s
semantics of modal sentences and also his metaphysics of possible worlds.
Because of isolatedness, Lewis’s theory of identity across worlds is not
based on strict identity, but rather on a specific similarity relation—counter-
parthood—to be discussed shortly. Furthermore, it is important to stress that
isolatedness concerns individuals, not properties or events. With the exception
of certain very special properties or events, for Lewis the same property is
exemplified in multiple worlds, and the same event occurs in multiple worlds.

9781472524263_txt_print.indd 96 01/12/2015 09:44

 MODAL REALISM 97

To determine whether individuals inhabiting different worlds are similar is to

compare the properties that those individuals share; and causation is defined
by tracking the same (sequence of) event(s) across multiple worlds, where
that (sequence of) event(s) is embedded in relevantly similar sequences of
events. (For a discussion of isolatedness in the context of modal realism,
good starting points are Bricker (2006a), (2006b), and (2007).)
Finally, no world is metaphysically privileged with respect to the others.
This feature is typically cashed out in the thesis that actuality is indexical.
That is, from the point of view of each world, that world is the actual world.
Expressions such as “here”, “now”, or “I” are indexical: without the specifi-
cation of a context (an index) with respect to which they are employed, those
expressions convey no specific meaning. Thus, the sentence:

4 Foffo is eating fish with potatoes now,

when uttered at 8:15 a.m. on Friday February 21, 2014, means:

5 Foffo eats fish with potatoes at 8:15 a.m. on Friday February 21, 2014.

But, when (4) is uttered on Saturday February 22, 2014, it means:

6 Foffo eats fish and potatoes at 8:15 a.m. on Saturday February 22, 2014.

Analogously, for Lewis, the sentence:

7 In the actual world: Foffo eats fish with potatoes,

when read by us, inhabitants of a certain world—call it W1—means:

8 At W1: Foffo eats fish with potatoes.

Yet, when (7) is uttered by an inhabitant of a world other than ours—call it

W2—it means:

9 At W2: Foffo eats fish with potatoes.

Typically, we regard the actual world as privileged with respect to other

worlds (see, for example, van Inwagen (1986) for a defense of this idea).
Such a privilege has, among others, an epistemic foundation: we have privi-
leged access to what happens in the actual world, and not what happens in
most other worlds, because we are acquainted with only the actual world.
Analogously, we tend to privilege the present, or the past, with respect to
the future: for instance, as we saw in Chapter 1 when discussing the problem
of Future Contingents, we regard the present and the past as fixed, whereas
the future we regard as open. To Lewis, however, there is no such privilege.

9781472524263_txt_print.indd 97 01/12/2015 09:44

 98 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

§4.2.2. Plenitude and recombination

Like many other philosophers, Lewis endorses the so-called principle of
plenitude of possibility. An intuitive, but sketchy and rather uninformative,
way to express the principle is that there is all that there could be. To begin
refining this intuitive formulation, a modal realist could say that, for any way a
world could be, there is a world that is that way. This formulation nonetheless
falls short of addressing the key philosophical question at stake, namely:
Which scenarios are possible?
In order to answer the latter question, and thus specify which worlds
exist, Lewis resorts to the so-called principle of “Humean recombination”,
named after David Hume, chief opponent of necessary connections in
metaphysics. The principle rests on the modal intuition according to which
anything can coexist, or fail to coexist, with anything else, as long as the
coexisting individuals can fit the maximal external relation of a world; for
the modal realist, the intuition amounts to the belief that “patching together
parts of different possible worlds yields another possible world” (Lewis 1986:
87–8). There are all the worlds that there could be by recombining whatever
collection of individuals with whatever other collection of individuals. The
individuals to be recombined need not belong to the same world; and neither
the individuals’ size nor their mereological complexity (i.e. how many parts
and how they are arranged) make a difference. We can thus offer a prelim-
inary formulation of the principle of recombination:

PR: Any collection of individuals x1, …, xn could coexist, or fail to coexist,

with any collection of individuals y1, …, yn, as long as they fit the
maximal external relation of a world.

PR, however, is unsuitable for a modal realist. This is because the modal
realist denies that any individual can inhabit more than one world, thus she
cannot endorse a principle according to which anything can coexist with
anything else. The modal realist must therefore translate PR into the language
of counterpart theory.
Before we can arrive at a modal realist formulation of PR, we must
consider parts of Lewis’s semantics for modal sentences. Counterpart
relations are relations of qualitative similarity among individuals; the range
of properties upon which the similarity rests varies based on the sentence
under consideration and the specific context in which it is used; thus, at a
minimum the similarity will rest on one property and, at a maximum, it will
rest on all the properties of the individuals under comparison. The flexibility
of counterpart relations is a theoretical virtue when it comes to explaining the

9781472524263_txt_print.indd 98 01/12/2015 09:44

 MODAL REALISM 99

meaning of sentences embedding modal expressions; however, when the

goal is to offer a formulation of the principle of recombination, such flexibility
poses a problem: we are not just recombining an initial cluster of individuals
(say, those of the actual world); rather, we have to devise worlds made out
of counterparts of the individuals in the initial cluster, and such operation
is more articulated because counterpart relation differs significantly from
identity.
Lewis solves the problem of offering a modal realist formulation of PR by
resorting to intrinsic properties and, through them, to the notion of an intrinsic
duplicate. In contemporary metaphysics, the debate regarding intrinsic
properties is complex and still open-ended (for a start, see Weatherson and
Marshall (2012)). Simplifying a bit, intrinsic duplicates are those individuals
that share all of their “perfectly natural” properties. Those properties have
two features: first, each of them serves to explain fundamental causal
relationships or to ground the similarity of natural kinds; second, each of
them is independent of accompaniment: each belongs to an individual
independently of whether that individual is, or is not, accompanied by any
other collection of individuals. Perfectly natural properties comprise the core
class of intrinsic properties: any other property that is intrinsic is defined
as such based on its relationship to the perfectly natural properties. More
precisely, the intrinsic properties of an individual are the properties that it
shares with all and only its intrinsic duplicates. Intrinsic properties are thus
independent of accompaniment; but, except for perfectly natural properties,
intrinsic properties cannot play a major role in the natural sciences (because
they do not carve nature at its joints as much as the perfectly natural
properties).
Because intrinsic properties are independent of accompaniment, they will
serve well the purpose of offering a modal realist formulation of the principle
of recombination, which may be stated as follows (for a more in-depth
discussion of the formulation, see Darby and Watson (2010)):

PR* For any collection of individuals x1, …, xn that stand in a maximal

external relation, there is a collection of individuals y1, …, yn such that:
(i) each of the y1, …, yn is numerically distinct from each of the x1, …,
xn; (ii) y1 is an intrinsic duplicate of x1, …, yn is an intrinsic duplicate
of xn; (iii) the y1, …, yn compose a world.

PR* expresses the thesis of the plenitude of possibility, according to the

modal realist.
Before ending this section, we should recall that the principle of recom-
bination—in its Humean or non-Humean variants—has been endorsed by
other philosophers. Most notoriously, the so-called combinatorialist theory of

9781472524263_txt_print.indd 99 01/12/2015 09:44

 100 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

possibility rests entirely on a certain understanding of the principle of recom-

bination. We will discuss this theory in Chapter 5, especially in connection
with the work of Brian Skirms and David Armstrong.

§4.3. Counterparts
In 1968, Lewis published a now-classic article on the semantics of sentences
containing modal expressions: “Counterpart Theory and Quantified Modal
Logic”. The article presented a novel modal semantics, which paved the
road for modal realism. Before examining this semantics in more detail,
and to place Lewis’s proposal in the context of this volume, it is important
to stress three characteristics of counterpart theory. First of all, the theory
moves within the framework of possible-worlds semantics, thus analyzing
the meaning of modal sentences in terms of sentences making reference to
possible worlds; by now, it should be clear that Lewis’s semantics, as well
as his metaphysics and epistemology, belong to the large group of theories
that analyze modal expressions in terms of possible worlds. Second—and we
have stressed that this is the element of novelty within Lewis’s semantics—
counterpart theory assumes that there is no overlap between the domains of
each world. (It is worthwhile to note, on a side, that the contention that no
world is metaphysically privileged with respect to another was not yet in place
in Lewis’s 1968 article on counterpart theory. Additionally, only later, when
Lewis fully developed his modal realism, was this second assumption refined
to the claim that worlds are isolated maximal individuals.) Third, the meaning
of modal sentences, for Lewis, is analyzed in terms of the relationship of
counterparthood.
Counterpart relationships, as we have already anticipated, are much more
flexible with respect to relationships of strict identity (see Lewis (1968),
Merricks (1999), and Borghini (2005)). To understand why, it is best to first
define counterparthood:

COUNTERPARTHOOD: An individual x1, inhabiting a world W1, is a

counterpart in W1 of an individual x2 inhabiting the world W2 when,
with respect to a given selection of properties P1, …, Pn, there is no
individual in W1 that resembles x2 more than x1.

Four salient characteristics of counterparthood tell it apart from identity (for

a broader understanding of counterparthood, see: Forbes (1982); Mondadari
(1983); Heller (2002, 2005); Dorr (2005); Fara and Williamson (2005); and Torza
(2011)).

9781472524263_txt_print.indd 100 01/12/2015 09:44

 MODAL REALISM 101

(i) Unlike strict identity, counterparthood is compatible with qualitative

dissimilarity and, thus, with an individual having more than one
counterpart in a given world. For instance, Foffo, the cat, may have
two cats as counterparts in another world, one having a second tail
that is grey and the other having a second tail that is black; both cats
are otherwise exactly like Foffo and more than anything else in their
world resemble Foffo.

(ii) Like strict identity, counterparthood is reflexive: any individual is

a counterpart of itself in its own world; it is not transitive: if x1 is a
counterpart of x2 in W1, and if x2 is a counterpart of x3 in W2, it is not
necessarily the case that x1 is a counterpart of x3 in W1.

(iii) Unlike strict identity, counterparthood is not symmetric: if x1 is

a counterpart of x2 in W1, it is not necessarily the case that x2 is a
counterpart of x1 in W2; there might be some x3 in W2 that resembles
x1 more than x2 resembles x1.

(iv) Finally, unlike strict identity, counterparthood is contextually defined

on the basis of a varying list of properties, broadly understood.

An examination of counterpart theory shows why properties play a key role

in Lewis’s metaphysics. It is by means of properties that the identity of
individuals across worlds can be fixed. Worlds are inhabited by non-overlapping
collections of individuals; on the other hand, properties are defined as sets
of individuals: each of those sets includes members that span across
multiple, possibly infinite, worlds. Thus, once we have the individuals and the
properties, we also have a picture of the relationships of qualitative similarity
between individuals inhabiting different worlds. Modal sentences express
some of those similarities. It was on the basis of this metaphysical apparatus
that Lewis provided a brilliant analysis of counterfactual statements in his
Counterfactuals (1973).
To recap, for Lewis the sentence:

10 Foffo could have had milk and cookies

is not really about Foffo itself. Rather, it is about certain counterparts of Foffo,
because the counterpart theorist’s translation of (10) is:

11 There is a world, other than the actual, and relevantly similar to the
actual, where a counterpart of Foffo has milk and cookies.

In other words, when the modal realist explains Foffo’s possible meal through
the existence of a world in which Foffo has a different meal, she is not
speaking literally. If we take things literally, Foffo could have had a different

9781472524263_txt_print.indd 101 01/12/2015 09:44

 102 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

meal seeing as there is a world in which an individual relevantly similar to

the one whom in our world we call “Foffo” has such a meal. This aspect
of the theory has been criticized time and again, generating the so-called
Humphrey objection, to be discussed below (see Kripke (1980), Merricks
(2003), and Sider (2006)).
We will end this section with a remark concerning the interplay of
counterpart theory and Lewis’s theory of individuals. For Lewis, there may be
individuals that, so to speak, are not world-bound. For instance, consider the
scattered whole composed of Foffo and of one of its counterparts in another
world; this whole is an individual that has parts in two distinct worlds. Modal
sentences from ordinary language, however, are never about such individuals.
(Of course, we can generate modal sentences about counterparts, possible
worlds, and so on, but those sentences belong to the language of the theory,
not to natural language. More on this later, when we discuss so-called
advanced modalizing.) Modal sentences are always about worlds or their
parts. (On this point, see especially Varzi (2001).)

§4.4. Modal realism and the philosophical

analysis of modality
Let us take stock. We have seen some fundamental metaphysical and
semantic tenets of modal realism, according to which:

(i) There are infinite worlds, as many as can be recombined by means of

PR*.

(ii) No individual inhabits more than one world.

(iii) Sentences of natural language containing modal expressions are
disguised qualitative comparisons of individuals (typically spread
across multiple worlds), according to the relation of counterparthood.

Based on (i)–(iii), we can now more precisely reformulate the solution to the
PP proposed by the modal realist:

MR*: The sentence of natural language: “It is possible that: s” expresses

the existence of some worlds W1, …, Wn in which “s*”,

where “s*” is a sentence obtained from s by substituting all of the individuals

in s with their counterparts in W1, …, Wn, based on the translation schemas
provided by counterpart theory. (In the last section of this chapter, we shall

9781472524263_txt_print.indd 102 01/12/2015 09:44

 MODAL REALISM 103

see that the translation schema must be flexible enough to accommodate

so-called advanced modalizing, that is, modal sentences that do not belong to
natural language, but rather to the language of some theory—most typically,
modal realism itself.)
Suppose, for the sake of the argument, that Lewis’s metaphysical and
semantic accounts of modality are fine and sound. A host of important episte-
mological questions would remain to be addressed. How can one prove, or
disprove, that Lewis’s picture is correct? How are we to know what is going
on in other worlds? After all, how do we even know that those worlds exist?
Lewis himself addressed these questions (cf. 1986: §2.4.). Certainly, it is
not through direct experience that we know of other worlds. We cannot travel
to them. Neither is it through imagination. Some philosophers have argued
that the worlds that exist are all and only those which we can imagine (see
the collection of essays in Gendler and O’Leary-Hawthorne (2002)). But a
modal realist is not committed to this view, and may have reason to reject
it. Perhaps there exist worlds that we cannot imagine. Or perhaps we can
imagine having a more developed imaginative capacity than the one we have;
if this is so, then there is also a world in which we have a more developed
imaginative capacity; but, then, if our world counts just as much as the other
in which inhabitants have more of an imagination than we do, why not believe
that there are worlds that we cannot imagine, but that our possible counter-
parts can? Furthermore, it is not necessarily true that all we can imagine
exists. Lewis himself offers the following admittedly controversial example:
we can imagine a square circle, but this does not mean that there is a world
with a square circle. See Lewis (1986: 113–15).
We will make some progress if we start by pointing out that Lewis’s
answer to the question “How many worlds are there?” rests on a modal
intuition, namely the intuition that anything can coexist with anything. Is it
then by means of a modal intuition that we are supposed to prove that modal
realism is true? Not quite, but that is the beginning. The rest of the argument
would depend on the plausibility of the philosophical analysis of modality
suggested by modal realism, and in particular the underlying conception of
philosophical analysis itself.
The idea that Lewis’s modal views should be understood in the light of,
and ultimately evaluated on the strength of, his conception of philosophical
analysis has been defended to a good extent by John Divers (see especially
(2013, 2014c)). To illustrate Lewis’s view of philosophical analysis, it is useful
to start with an often quoted passage from Counterfactuals:

One comes to philosophy already endowed with a stock of opinions. It

is not the business of philosophy either to undermine or to justify these
preexisting opinions, to any great extent, but only to try to discover ways of

9781472524263_txt_print.indd 103 01/12/2015 09:44

 104 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

expanding them into an orderly system. A metaphysician’s analysis of mind

is an attempt at systematizing our opinions about mind. It succeeds to the
extent that (1) it is systematic, and (2) it respects those of our pre-philo-
sophical opinions to which we are firmly attached. Insofar as it does both
better than any alternative we have thought of, we give it credence. There is
some give-and-take, but not too much […] So it is throughout metaphysics;
and so it is with my doctrine of realism about possible worlds. Among my
common opinions that philosophy must respect (if it is to deserve credence)
are not only my naïve belief in tables and chairs, but also my naïve belief
that these tables and chairs might have been otherwise arranged. Realism
about possible worlds is an attempt, the only successful attempt I know of,
to systematize these preexisting modal opinions. (1973: 88)

We can derive two important lessons. The first is that we should not think of
philosophical analysis as a philosophical process that solves once and for all
the questions we ask regarding a certain topic. Thus, our analysis of modal
expressions is not to be regarded as successful just in case it leaves us
with no more questions regarding the meaning of such expressions. Lewis
confirms this lesson just a few lines after the quoted passage:

But why should I think that I ought to be able to make up my mind on every
question about possible worlds, when it seems clear that I may have no
way whatever of finding out the answers to other questions about noncon-
tingent matters—for instance, about the infinite cardinals?

The second lesson regards the philosophical method suggested by Lewis’s

analysis. We can divide it into three parts: Opinion, Analytic Hypothesis, and
Metaphysical Basis. The Opinion is the set of naïve beliefs with which we
come to philosophize about a subject matter. As we have seen, a crucial naïve
belief guiding our philosophical theorizing about possibility and necessity is
the naïve formulation of the principle of recombination (i.e. that anything can
coexist with anything). The Analytic Hypothesis is the core contribution of
a theory, defining the new terms introduced by the theory and thus speci-
fying how they should be understood (and employed). (Typically, the Analytic
Hypothesis is expressed in terms of so-called Ramsey sentences, where the
empirical content of the newly introduced theoretical term is made explicit.
We will not get into the details of this here, though.) Finally, the Metaphysical
Basis makes explicit the ontological commitments of the theory, that is, what
must exist—if anything—in order for the theory to be true.
Equipped with an outline of Lewis’s method of philosophical analysis, we
can now summarize his analyses of modal plenitude and of possibility and
necessity. Let us begin with the analysis of plenitude:

9781472524263_txt_print.indd 104 01/12/2015 09:44

 MODAL REALISM 105

(Opinion) There is all that there could be. Anything can coexist with
anything else.
(Analytic Hypothesis) PR*: For any collection of individuals x1, …, xn that
stand in a maximal external relation, there is a collection of individuals
y1, …, yn such that: (i) each of the y1, …, yn is numerically distinct from
each of the x1, …, xn; (ii) y1 is an intrinsic duplicate of x1, …, yn is an
intrinsic duplicate of xn; (iii) the y1, …, yn compose a world.
(Metaphysical Basis) An infinite stock of individuals (including infinite
maximal individuals), properties, and relations (including infinite
external relations) exists.

Notice that the Metaphysical Basis countenances the possibility that worlds
contain properties completely different from those exemplified by individuals
in our world—so different that these properties are alien to our world (the
expression is employed by Lewis himself). So, as far as we know, the actual
world may not suffice in providing the metaphysical basis for the theory.
As we shall see, Lewis regarded this point as a virtue of his theory, shared
by none of the other major rival theories of modality employing possible
worlds (with the exception, perhaps, of fictionalism, as we will see in
Chapter 6). We shall come back to alien properties in Chapter 5, when
discussing Combinatorialism.
Let us now turn to the philosophical analysis of possibility and necessity.

(Opinion) Most situations we experience are contingent. For instance,

Pierluigi might have been with me right now; and I could have avoided
writing this sentence.
(Analytic Hypothesis) For any sentence P, P is possible if and only if
P’s translation in the language of counterpart theory—call it P*—,
understood in accordance with the metaphysical tenets of modal
realism, is true.
(Metaphysical Basis) An infinite stock of individuals (including infinite
maximal individuals), properties, and relations (including infinite
external relations) exists.

The Metaphysical Basis of the analysis of possibility and necessity is the

same as the Metaphysical Basis of the analysis of modal plenitude. The
Analytic Hypotheses, however, reveal that two different bits of philosophical
theorizing are employed to analyze different sets of naïve opinions.
The analyses just provided suggest the direction taken by Lewis in
addressing the EPP: we know of the existence of other worlds, and we know
what those worlds are like, because the best philosophical theory we have
at hand says that they exist. It is not a matter of acquaintance. It is not a

9781472524263_txt_print.indd 105 01/12/2015 09:44

 106 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

matter of imagination. It is, rather, a matter of philosophical understanding.

Such understanding is not an all-or-nothing affair, but rather expresses itself
in a complex system of tenets, some of which may be uncontroversial, while
about others we may argue back and forth; furthermore, philosophical under-
standing does not answer all our questions regarding the subject matter that
our theories systematize. Lewis in effect proposes a rationalistic model of the
epistemic justification of a theory of modality that is more modest and more
open to revision than rival models. For example, recall Lowe’s justification of
essentialism offered in Chapter 3: there, the analysis of modality was justified
by appealing to allegedly uncontroversial intuitions regarding real definitions
of individuals. The analysis left little room for revisions and for dialogue
between theorizing and naïve opinions. To date, the distinctive nature of
Lewis’s contribution to modal epistemology has been recognized, but it is yet
to be studied and appreciated in more detail.

§4.5. The theoretical advantages of

modal realism
Let us go back to the passage cited at the beginning of this chapter, where
Lewis explains that we should believe in a plurality of worlds because of the
ensuing theoretical benefits of such a choice. It is now time to discuss those
benefits in more detail. We will outline five of them, drawn, by and large, from
Chapter 1 of Lewis (1986).
The first theoretical benefit of modal realism is that, unlike any other
theory of possibility, it aims to reduce modal entities to non-modal entities.
That which is possible is that which exists at a certain world (the charac-
teristics of the relevant world depend on the specific sentence expressing
the possibility). That which is necessary is that which exists at all worlds.
(Or, most often, at all worlds of a certain kind, as a tacit implicit restriction
is typically embedded in a sentence expressing a necessary truth.) Modal
realism thus provides an analysis of possibility and necessity in terms of
existence. (Incidentally, we may notice that it is for this reason that Lewis’s
theory was almost immediately regarded as suspiciously Meinongian: Lewis
seems to endorse the existence of certain things—namely, things involved
in merely possible scenarios—that—it seems fair to say—do not exist,
since they do not actually exist. We shall come back to Meinongianism in
Chapter 8.) Although the reduction of modality to simple existence seems at
first glance to be an unremarkable result, it should be noted that none of the
other theorists discussed in this volume even considered this achievement
within reach. To a modal realist, speaking of capacity and probability, making

9781472524263_txt_print.indd 106 01/12/2015 09:44

 MODAL REALISM 107

hypotheses, speaking of the future, and so on, amount to speaking of

that which exists. In fact, from the modal realist’s theoretical perspective,
unicorns exist as much as mountain gorillas: for, surely there is a world where
there are unicorns, because there seems to be no contradiction in the idea of
a unicorn; and surely mountain gorillas are still around on Earth, even if only
a few hundred are left. Some, like the expressionist and the skeptic, believe
that possibility is merely a projection of—respectively—human sentiments
or something confusing and unfathomable; others—namely the modalist and
other theorists we have yet to consider—believe that what is possible exists
in its own way. The modal realist is the only one who can say that that which
is possible is simply no different from that which is actual, and who can
interpret what we say about that which is possible in the same way as what
we say about that which is actual.
The second theoretical benefit of modal realism is that it yields a particu-
larly simple analysis of so-called counterfactuals, sentences that express a
hypothetical scenario that is contrary to an actual scenario, for example:

12 If I hadn’t been here writing this book, I would have gone surfing.

Counterfactuals pose an obvious problem for those who do not accept

the existence of non-actual entities, as it seems that that which makes a
counterfactual sentence true cannot be rooted in actuality (though this was
recently called into question by Williamson (2007a, 2007b, 2010). Sentence
(12) expresses a mere possibility: a scenario which is not (and was not and
will not be) realized in the actual world; so, the antecedent of (12) is false in
the actual world. In classical logic, a sentence whose antecedent is false is
always true. But this is not the case with all counterfactuals. For instance,
while I may be inclined to think that (12) is true, I may be inclined to regard
the following sentence as false:

13 If I hadn’t been here writing this book, I would have gone to rob a bank.

Despite the fact that (12) and (13) have the same antecedent (“If I hadn’t been
here writing this book”), and that this antecedent is false in both (12) and (13),
my inclination is to endorse (12) and reject (13). But how to go about arguing
for my intuitions? The modal realist offers an elegant philosophical analysis of
counterfactuals that answers this question. According to Lewis, (12) and (13)
should be analyzed as follows:

14 There is a world, and a counterpart of mine in that world, where that

counterpart decided to go surfing instead of writing for the day; and
that world, compared to all the other worlds in which my counterparts
decide not to write today, is the most similar to the actual world.

9781472524263_txt_print.indd 107 01/12/2015 09:44

 108 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

15 There is a world, and a counterpart of mine in that world, where

that counterpart decided to go rob a bank instead of writing for the
day; and that world, compared to all the other worlds in which my
counterparts decide not to write today, is the most similar to the
actual world.

While the concept of resemblance, embedded in counterpart relations, is

vague and maybe a bit elusive, the Lewisian solution offers a straightforward
explanation of the contents of (12) and (13), clarifying why I may be inclined
to believe (12), but to disbelieve (13): were everything to stay the same except
that I would not have been writing, it seems highly implausible to me that I
would have gone to rob a bank today, while it seems highly plausible that
I would have gone surfing.
The third theoretical benefit of modal realism rests in the metaphysical
robustness of the theory. According to modal realism, there are many more
possibilities than the ones we can express in our language. We have the
actual world; we have all worlds we can imagine, as long as they are not
contradictory; we have all worlds that our languages can represent, as long as
there is no contradiction in such representations; and we have any additional
world that there could be. Modal realism, in other words, is compatible with
a sort of metaphysical humility, according to which the actual world does
not contain all the kinds of things that there could be: there may be many
more kinds of things than there are in this world and that we can express or
imagine. No other theory can defend such a claim, according to Lewis (see
especially (1986: Chapter 3)); for this reason, modal realism is the metaphysi-
cally most robust theory on the market.
The fourth theoretical benefit is that modal realism calls for an ontology
that is, in a certain sense, minimal. All there is are worlds (and their parts)
and sets of worlds or their parts. The modal realist might even be able to do
without these sets (see Lewis (1991)), but I will not get into the details of
this proposal here. Through the worlds, their parts, and the sets of worlds
and parts, we are able to define any other ontological category. A property is
a set of parts of worlds; for example, redness is the set of all of the things
that are red in all of the worlds. A proposition (the content of a sentence) is a
set of worlds, namely, those where it is true (with the exception of so-called
de se propositions—see Lewis (1979)). The so-called “natural kinds” (roughly
speaking: the properties that play a key role in scientific laws) would be
certain properties, that is, certain sets. And so on, and so forth.
The fifth and last theoretical benefit of modal realism is that it offers a
simple and elegant model to explain the behavior of agents. Before acting,
each one of us contemplates various propositions and then, usually, chooses
to actualize one of these propositions. Each proposition corresponds to a set

9781472524263_txt_print.indd 108 01/12/2015 09:44

 MODAL REALISM 109

of worlds, as mentioned. Choosing a proposition on which to act amounts

to undertaking that our world is the one in which that proposition will be
realized. If I desire something that cannot be realized, it is most likely because
I am contemplating a proposition that contradicts another proposition that is
already true of my world. This type of model allows us to form intuitive repre-
sentations—some highly complex—of the behavior of an agent. See Lewis
(1986), Stalnaker (1984), and Stalnaker (2003).
The Lewisian picture of modality remains one of the most powerful
and elegant philosophical theories of modality put forward: it rests on an
innovative and flexible semantic theory; it is supported by a robust and
elegant metaphysical view; and it is backed by a plausible epistemological
story relying on philosophical analysis. And yet, readers have doubted that it
really possesses the five theoretical benefits suggested by Lewis, at times
harshly criticizing his arguments aimed at demonstrating such benefits, so
much so that today there are a considerable number of apparently good
reasons to distance ourselves from the Lewisian perspective. A selection of
these will be considered in the next section.

§4.6. Modal realism and the definition of

modal concepts
The criticisms made of modal realism are numerous, and most of them have
generated considerable debate, making it difficult to provide a thorough
overview. Lewis (1986: Chapter 2) addresses eight lines of objection, and
over the years a few more have been added. The reader will find additional
references to related literature in the Further Reading section. In this
section, we shall concentrate on a particularly central criticism of Lewis’s
theory, according to which modal realism is unable to define—and therefore
eliminate—modal concepts; since the reduction of modal concepts is
regarded as the major theoretical prize for endorsing the bizarre idea that
there are infinite worlds on a par with ours, the criticism threatens to cast
serious doubt on the plausibility of the philosophical analysis proposed by
Lewis.
There are different ways to rehearse the criticism. One that seems particu-
larly straightforward begins with the principle of recombination. We shall start
by recalling the final version of the principle we provided earlier:

PR* For any collection of individuals x1, …, xn that stand in a maximal

external relation, there is a collection of individuals y1, …, yn such that:
(i) each of the y1, …, yn is numerically distinct from each of the x1, …,

9781472524263_txt_print.indd 109 01/12/2015 09:44

 110 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

xn; (ii) y1 is an intrinsic duplicate of x1, …, yn is an intrinsic duplicate

of xn; (iii) the y1, …, yn compose a world.

Principle PR* employs modal concepts in two ways. To see why this is the
case, let us again unfold the reasoning at the basis of the principle of recom-
bination. The goal of PR* is to spell out the modal intuition that anything can
coexist, or fail to coexist, with anything in terms of a non-modal metaphysical
principle. We can break down the achievement of this goal into four steps.
The first step is to replace the “anything” in the modal intuition with collec-
tions of individuals. The second step is to exploit the modal realist’s thesis that
there are infinite worlds, thereby reducing the “can coexist” to the non-modal
coexistence of counterparts in some world. The third step is to qualify those
counterparts as intrinsic duplicates. The fourth and final step is to ensure that
the intrinsic duplicates can fit a certain reticulate of external relations.
Modalities surface at the third and fourth step. Let us begin with the latter.
According to the criticism under consideration, the appeal to the possibility
that the intrinsic duplicates fit the reticulate of external relations introduces
a metaphysical modality into PR*. It is a metaphysical modality because it
countenances the possible coexistence of the intrinsic duplicates and of a
certain system of external relations. So, here is the problem. The modal realist
is supposed to analyze the possible ways that individuals could be in terms
of the existence of counterparts of those individuals; PR* is the principle that
should clarify which individuals exist, unrestrictedly, so that we generate
enough counterparts to support the theory; PR* appeals to the possible
coexistence of some individuals and certain reticulates of external relations
among them; this appeal should be explained in terms of counterparts;
thus, the very principle that is supposed to tell us what are the individuals in
the theory that support the appeal to counterpart relations appeals to such
relations; PR* fails to provide an adequate metaphysical basis for the theory
because it cannot provide an adequate stock of individuals.
The modal realist may reply that the appeal to the possible coexistence is
a case of advanced modalizing and, thus, does not require the existence of
counterparts to have been postulated. Advanced modalizing is an important
part of modal realism. Consider the sentence:

16 There could have been more worlds than there are.

Sentence (16) cannot be analyzed in terms of counterpart relations, because

this would imply that there are more worlds than in fact there are, which is
impossible. (Incidentally: it is on the heels of sentences such as (16) that
the debate on impossible worlds ensued, starting especially with Yagisawa
(1988), on the heels of Lewis (1986, §3.2); we shall return to these issues

9781472524263_txt_print.indd 110 01/12/2015 09:44

 MODAL REALISM 111

in Chapter 8, where we will discuss dialetheism.) Instead, the modal realist

argues that sentences such as (16) are part of so-called advanced modalizing:
modal talk that concerns theoretical entities and that therefore does not really
involve natural language, but the language of the theory we have put forward
to deal with modalities in natural language. The aim of modal realism is to
provide a reductive analysis of modal talk in natural language; if the analysis
cannot apply to theoretical language, that is fine: for theoretical language
we can introduce a sui generis analysis that makes no appeal to counterpart
relations. For example, the modal realist would insist that (16) be read as:

17 There are more worlds than there are,

which is patently contradictory and accords with our intuition that (16) affirms
the existence of impossible worlds. On analogy with the suggested reading of
(16), the modal realist could insist that the modality in the fourth step of PR* is
an instance of advanced modalizing and thus involves no counterpart relation.
In PR*, the appeal to an “appropriate” maximal external relation should be
understood as meaning something like “matching”—where the match in
question suggests instructions for how the worlds should be theoretically
analyzed—and not as introducing a genuine metaphysical modality.
We shall now consider the second count on which modal realism may
be accused of failing at reducing modalities. As we indicated above, this
objection rests on the appeal to intrinsic duplicates in PR*. Intrinsic dupli-
cates, we shall recall, are defined by Lewis in terms of natural properties,
that is, those properties that supposedly ground fundamental scientific claims
and that are independent of accompaniment (the individual possesses that
property regardless of whether any other individual exists). The appeal to
independence of accompaniment, however, involves yet again a modality. A
good way to explain this is to point out, as have Cameron (2008) and Denby
(2008), that the definition of intrinsic duplication implicitly presupposes a
principle of recombination; in fact, independence of accompaniment means
that a given individual is combined with other individuals in every possible
way. Thus, PR* is spelled out in terms of intrinsic duplicates; and the latter are
spelled out in terms of PR*. What we are left with is plain circularity, which
can be illustrated in the following schema:

Modality
⇓
Existence of infinite worlds
⇓
Principle of Recombination
⇓

9781472524263_txt_print.indd 111 01/12/2015 09:44

 112 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Intrinsic Duplication
⇓
Principle of Recombination

Certainly, a more modest formulation of the principle of recombination that

appeals to generic counterparts, and not intrinsic duplicates, would avoid the
circularity:

PRC: For any collection of individuals x1, …, xn that stand in an

appropriate maximal external relation, there is a collection of
individuals y1, …, yn such that: (i) each of the y1, …, yn is numerically
distinct from each of the x1, …, xn; (ii) y1 is a counterpart of x1, …, yn is
a counterpart of xn; (iii) the y1, …, yn compose a world.

But PRC has too high of a price, given the aims of the modal realist. In fact,
PRC greatly limits the number of possibilities that the theory can represent,
given that counterpart relations often take into account the environmental
circumstances of an individual and, thus, are not independent of accompa-
niment. Modal realism, then, would no longer boast a greater number of
possibilities than its rival theoretical proposals.
Is Lewis’s modal realism a theoretical paradise for scholars of modality? As
we suggested, Lewis’s proposal offers simple and elegant solutions to the PP,
MPP, and EPP. Probably, from this perspective, it offers the best integrated
and most solid theory on the market. And yet, modal realism has important
drawbacks, and not only because it compels us to accept a bizarre account
of worlds. As we have seen in the current section, modal realism has some
important theoretical limitations. And we have just scratched the surface. For
lack of space we have not addressed other potential sources of theoretical
concern: for instance, the tenability of the Lewisian definition of actuality
as indexical, of cases of recombination in connection with the possibility of
so-called gunk (i.e. a whole with infinite mereological complexity), and of the
number of worlds. In Chapter 8, we shall return to some of these topics. Now,
we shall turn to the main rival of modal realism, ersatzism.

Study Questions
• What is the modal realist’s reply to the MPP?
• What is the principle of recombination?
• What is the modal realist’s version of the principle of recombination?

9781472524263_txt_print.indd 112 01/12/2015 09:44

 MODAL REALISM 113

• What is a natural property, according to Lewis?

• What is an intrinsic duplicate, according to Lewis?
• What is the modal realist’s solution to the SPP?
• What is a counterpart relation?
• What is the modal realist’s solution to the EPP?
• For Lewis, philosophical analysis involves three steps: what are they?
• What are the five main theoretical virtues of modal realism?
• Why might the principle of recombination conceal some modalities?
• What is advanced modalizing?

FURTHER READING
For an overview of modal realism, we recommend starting from Lewis’s writings,
most especially Lewis (1973: 84–91, 1986: Chapter 1). For a shorter but still detailed
introduction, see especially Divers (2002: Part II) and also Melia (2003: Chapter 5).
The latter two texts also illustrate, or contain bibliographical references on, specific
aspects of modal realism and counterpart theory. Among the overviews of modal
realism, the reader may also consult Chihara (1998: Chapter 3).
Particularly useful for framing modal realism within the broader context of
Lewis’s philosophical work are the following volumes on the philosophy of
Lewis: the collection of essays edited by Preyer and Siebelt (2001), especially the
first part, and also those edited by Jackson and Priest (2004), in particular Chapters
3, 12, and 20; Daniel Nolan’s introduction (2005); the companion devoted to the
philosophy of Lewis, edited by Schaffer and Loewer (2015); and the special issue
of Humana.Mente edited by Carrara, Ciuni, and Lando (2011).
The original presentation of counterpart theory is Lewis (1968), which should
be read along with its postscript, published in Lewis (1983), and integrated with
the consultation of Forbes (1982; also 1985: Chapter 3) and Mondadori (1983).
Counterpart theory has been criticized on various counts—to start see: Merricks
(1999, 2003); Heller (2002, 2005); Dorr, Fara, and Williamson, respectively (all
2005); Borghini (2005); and Sider (2006). Additional criticisms are addressed by
Lewis in his work of 1986 (see Chapter 2).
Unusually vast is the literature that critically assesses one or another aspect
of Lewis’s metaphysics. John Divers has made an important contribution for
understanding the modal status of modal realism, as well as Lewis’s method of
philosophical analysis; see especially Divers (2012, 2013, 2014a, 2014b, 2014c). As
early as 1979, William Lycan claimed that the idea of a (possible) world is in itself
modal (cf. 1979). Shalkowski (1994) reiterated the doubt about whether modal
realism is effective in reducing modalities to the existence of individuals at worlds.
The doubts are rehearsed in Divers (2002: Part II), Melia (2003: Chapter 5). For a
broader discussion of the metaphysical aspects of modal realism, see also work by
▲

9781472524263_txt_print.indd 113 01/12/2015 09:44

 114 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Philip Bricker (in particular, (2006a, 2006b, 2007)), and by Kris McDaniel (2006). For
a recent discussion of the principle of recombination, see Denby (2008), Cameron
(2008), and Darby and Watson (2010). Finally, Peter van Inwagen has offered some
insightful contributions to the debate on modal realism (especially 1980, 1986).

9781472524263_txt_print.indd 114 01/12/2015 09:44

 5
Ersatzism

E rsatzism counts the largest number of subscribers among the theories of

modality. For this reason, it is difficult to provide a taxonomy of the ersatzist
positions. All ersatzist positions share the conviction that possible worlds are a
surrogate (in German: ersatz) of the actual world. The positions differ with respect
to the metaphysical status of the surrogates. After a brief historical and general
presentation of ersatzism (§5.1.), four positions are considered, and critically
assessed: linguistic ersatzism (§5.2.), combinatorialism (§5.3.), pictorial ersatzism
(§5.4.), and atomic ersatzism (§5.5.). The taxonomy, by and large, follows the one
adopted by Lewis (1986), with the addition of combinatorialism.

*****

§5.1. Introduction

L et us suppose, together with the modal realist and the modalist, that
possible-worlds semantics is the best instrument available for addressing
the numerous, complex conceptual nexuses related to the PP; let us also
assume, together with the modal realist, that possible worlds can be philo-
sophically analyzed, thus delivering a philosophically savvy interpretation
of the diamond and box (the symbols expressing, respectively, possibility
and necessity in the language of quantified modal logic (QML)). Must we,
therefore, endorse the theory, advocated by the modal realist, that there are
infinite worlds like ours? No. In fact, the majority of those who would explain

9781472524263_txt_print.indd 115 01/12/2015 09:44

 116 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

the meaning of modal sentences in the terms of a philosophical theory of

possible worlds believe that those worlds are surrogates for certain elements
of the actual world. This is why their position takes the name ersatzism, from
the German ersatz: “surrogate” or “substitute”. Normally included among the
supporters of this theory is Leibniz, the figure who introduced the concept of
a possible world in modern philosophy (see Chapter 1).
Of the theories of possibility presented in this book, ersatzism is perhaps
the one that boasts the longest tradition. It underlies the positions of a wide
array of twentieth-century and contemporary authors, including Ludwig
Wittgenstein, Rudolf Carnap, Alvin Plantinga, Robert Stalnaker, and David
Armstrong, to cite just a few influential names. For this very reason, providing
a comprehensive overview of ersatzism that takes into account the positions
of its proponents with regard to the three sub-problems of possibility (SPP,
MPP, and EPP), and with regard to the nature of language, truth, knowledge,
existence, properties, and individuals, is an arduous task. In what follows,
we will take an in-depth look at the two positions that, throughout the years,
have become most prominent, namely linguistic ersatzism (§5.2.) and combi-
natorialism (§5.3.). We will then consider two other relevant positions, which
are also illustrated in Lewis (1986: Chapter 3): pictorial ersatzism (§5.4.) and
atomic ersatzism (§5.5.).
The traits differentiating the four positions under review are mainly
metaphysical and epistemic; at the semantic level, they all typically endorse
some version of QML, with identity of individuals across worlds (thus,
they reject counterpart theory). Linguistic ersatzism contends that possible
worlds are linguistic entities: each world is a maximally consistent (and
possibly infinite) set of sentences; combinatorialism, by contrast, maintains
that possible worlds are abstract recombinations of elements of the actual
world, where a recombination is not a linguistic entity, but an abstract entity
defined on the basis of the individuals, properties, and natural laws of the
actual world; pictorial ersatzism replies that possible worlds are, rather,
non-linguistic representations of the way the world could be (e.g. proposi-
tions or states of affairs that represent the way our world could be), so
that the number of possible scenarios is not tied to the expressive powers
of language; finally, atomic ersatzism proposes that possible worlds are
abstract entities devoid of structure, and thus simple, atomic. Though the four
positions are characterized by their metaphysical take on possible worlds, it is
important to stress that such metaphysical divergences often go along with
epistemic ones.
Ersatzists have been the most fierce opponents of modal realism. The key
points of rupture between the two positions are semantic and metaphysical.
At the semantic level, ersatzists object to counterpart theory, especially the
idea that there is no overlap between the domains of individuals of any two

9781472524263_txt_print.indd 116 01/12/2015 09:44

 ERSATZISM 117

worlds, endorsing instead the initial formulation of possible-worlds semantics,

in which the same individual inhabits multiple (possibly infinite) worlds. At the
metaphysical level, ersatzism affirms the priority of the actual world over all
other worlds, which are indeed regarded as surrogates of actuality.
Perhaps the most natural way to think of possibility is, in fact, to believe
that there is only one world: the actual world, in which we live. If, however,
we believe that the ultimate explanation of the meaning of sentences
containing modalities requires a theory of possible worlds, then we are faced
with the problem of having to explain where these worlds can be found in the
actual world. Hence the ersatzist solution to the MPP:

ME: Only one world exists, the actual world, of which possible worlds
are surrogates.

Surrogates, hence, stand in as replacements for the purposes of someone’s

speaking or thinking intelligibly of some possibility. Consider, for example, the
following sentence:

1 Foffo the cat could have had milk and cookies.

Even the ersatzist would assert that there is a strict relationship between (1)
and the following:

2 It is possible that: Foffo the cat has milk and cookies.

And (2), just as for the modal realist, is in turn explained in terms of the truth
of the following sentence:

3 There is a world where Foffo the cat has milk and cookies.

Yet, when it comes to clarifying the metaphysical commitments of (3),

ersatzism and modal realism diverge. According to the latter, as we have seen,
(3) commits us to the existence of a world that is on a par with our own and
where a counterpart of Foffo (not Foffo, the cat, itself, but an individual that
most resembles Foffo in that world) has milk and cookies; the ersatzist, on the
other hand, would argue that (3) commits us to the existence of surrogates of
entities of our world, including a surrogate of Foffo, the cat, and a surrogate
of the milk and cookies of our world. But what exactly is a surrogate?
The various ersatzist roads diverge exactly at this point. Building upon the
taxonomies adopted by Lewis (1986), Divers (2002), and Melia (2003), we
will follow four main ersatzist avenues. In doing so, we will encounter four
interpretations of the ontological allegiances of (3) and four sets of solutions
to the MPP and EPP.

9781472524263_txt_print.indd 117 01/12/2015 09:44

 118 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

§5.2. Linguistic ersatzism

The first approach considers a possible world to be a surrogate of a set of
sentences in a certain language. This position is known as linguistic ersatzism
and is a longstanding tradition in modal theorizing. Versions of linguistic
ersatzism can be attributed to some of the founders of possible-worlds
semantics, including Rudolf Carnap and Jaakko Hintikka.
The key idea of linguistic ersatzism is that, in the same way that the
sentence:

4 Pinocchio was carved by Geppetto

is true according to (or, in brief: at) a certain story, namely the book Pinocchio,
so the sentence:

5 Foffo could have had milk and cookies

is true according to (or, in brief: at) a certain world. Like a book, a world is a
very long—possibly infinitely long—and complex list of sentences; in fact,
a world is the set of those sentences.
More precisely, the linguistic ersatzist will define a world thus:

WLE: A possible world is a complete and consistent description, given in

a certain language, of a state of things.

A world in which Foffo has milk and cookies is nothing but a description of
a state of things—given in English, for example—in which Foffo enjoys milk
and cookies. As such, WLE also indicates linguistic ersatzism’s answer to the
MPP: in short, possible, non-actual entities are sets of sentences.
As descriptions, possible worlds are fundamentally differentiated from our
world by the fact that they are not concrete, in the way that we would say
our world is: our world is not a mere description; it is made of things, not
words. Thus, unlike modal realism, linguistic ersatzism denies that “actual” is
an indexical and that the actual world is metaphysically no more or less privi-
leged than any non-actual world; on the contrary, there is a major difference,
as possible worlds consist simply in sentences. It is at this point important
to point out the danger of confusing linguistic ersatzism and modal fiction-
alism, a theory we will discuss in Chapter 6. One could in fact think that, if
possible worlds are sets of sentences, just like books, then in a sense they
are linguistic fictions. This is fine, but it is far from implying that linguistic
ersatzism and modal fictionalism are one and the same theory. As we shall
see, modal fictionalism treats as fictions certain theories of modality, such

9781472524263_txt_print.indd 118 01/12/2015 09:44

 ERSATZISM 119

as modal realism: thus, the distinctive suggestion of modal fictionalism is—

say—that we should endorse modal realism, yet treat it as a useful fiction.
This is not what linguistic ersatzism advocates. So, even if linguistic ersatzism
and modal fictionalism ultimately share the claim that possible worlds are
linguistic entities, they differ in how such linguistic entities are generated:
for the modal fictionalist, possible worlds arise out of the fiction a theory of
modality. Let us now turn to the details of how possible worlds are generated
for the linguistic ersatzist.
We must first specify the requirements a set of sentences must fulfill in
order to count as a possible world. Afterward, we can see how such sets can
be generated. Two requirements are especially relevant: completeness and
consistency. First of all, the set should be detailed enough to include all that
there is to describe in a world. More precisely, a set of sentences describing
a world is maximal (or: complete) when, for every describable scenario of that
world, the set includes either a description of that scenario or a description
of its negation. Second, the set should be consistent. That is, it should not
contain two descriptions of scenarios such that one is the negation of the
other. According to some authors, the requirement of consistency may
be given up without compromising the spirit of linguistic ersatzism; for a
discussion of impossible worlds, including sets of sentences that are not
consistent, see Chapter 8.
By characterizing a possible world as a maximal and consistent set
of sentences, the linguistic ersatzist utilizes certain modal concepts—
completeness and consistency—in order to explain another modal
concept—possible worlds. Maximality is a modal concept because it appeals
to describable scenarios; consistency is a modal concept because it appeals to
the idea of incompatibility, which we expressed by the requirement that the
set of sentences could not contain two descriptions such that one is the
negation of the other. To the linguistic ersatzist, the modal circularity is not
a major reason for concern, though. It is not part of the general project
of ersatzism to reduce modal talk to non-modal talk, or to explain—as the
modal realist attempts to do—possible and necessary entities in terms of
the existence of certain non-modal entities. Linguistic ersatzism will succeed
if it can provide an adequate analysis of modal sentences in terms of sets of
sentences; insofar as the view provides answers to the MPP and EPP that
dissipate the doubts that lead us to formulate them, it does not matter
that those sets of sentences are themselves modal entities. So, the issue is:
does linguistic ersatzism provide satisfactory answers to the MPP and EPP?
When it comes to the MPP, linguistic ersatzism may have an advantage
over modal realism. For, unlike the latter, the former may more easily explain
what a (possible) world is. For instance, it may suffice to show that a language
exists independently of it being used, thereby showing also that (possible)

9781472524263_txt_print.indd 119 01/12/2015 09:44

 120 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

worlds exist independently of us talking about them. The point of contention,

however, is whether linguistic ersatzism is able to supply a sufficient amount
of worlds to satisfy our pre-theoretical intuitions about possibility. Lewis and
other authors have questioned linguistic ersatzism on precisely this point.
How does a linguistic ersatzist generate possible worlds? One way, of
course, would be to describe them in a natural language. But more systematic
strategies can be found. A well-known one builds upon a so-called Lagadonian
language (from the language experiment of the Academy of Science at Lagado,
described in Jonathan Swift’s Gulliver’s Travels). Suppose that every individual
a in a world names itself—a’s name being “a”—and that each property P and
relation R also is its own predicate—respectively, “P” and “R”. At this point,
a sentence that describes a world will be an n-tuple containing n–1 names of
individuals and a property or a relation; for example, the triple <R, a, b> will be
the sentence expressing the fact that the relation R holds between individuals
a and b. So, at a world w it is true that R holds between individuals a and b if
and only if <R, a, b> is included in the set of sentences that characterizes w.
If we believe that our assertions about possibility are endowed with
meaning, then we would presumably also believe that the totality of the
scenarios they may describe is infinite. Two apples could exist, or three, or …
It seems not contradictory for there to be infinitely many apples. Furthermore,
there could be apples, but also pears, oranges, apricots … It seems not
contradictory for there to be an infinite number of other fruits, infinitely many
of which did not, do not, and will not exist in our world. In order for the
description of a world to be complete, it must contain all describable states
of things. But here we encounter the difficulty of producing a number of
sentences that is equal to the number of possibilities. We encounter, that is,
the problem of generating worlds that are suitable to represent the scenarios
that intuitively we regard as possible.
In fact, as observed by Lewis (1986: §3.2.), there could be many more
worlds than those we are able to express through a language. Let us
suppose that space is continuous: that is to say, between any two points
in space another point can always be found. We might wish to express the
possibility that each of these points can be occupied or not occupied. But a
language is arguably unable to provide a description of a cardinality of states
of things equal to that of the states of the things that constitute a continuous
space; this is a straightforward result of our failure to prove the continuum
hypothesis, according to which—roughly—any set of real numbers can be
put into correspondence with a set of natural numbers. So, through language,
we have no way of expressing the existence of a world made of points that
occupy a continuous space.
The scenarios that are suggested by our modal intuitions can be even more
complex: besides space, time could be continuous; or we could have infinite

9781472524263_txt_print.indd 120 01/12/2015 09:44

 ERSATZISM 121

properties for each point. And this is all without taking into account that space
and time can be extended to infinity. A natural language is far from being
able to express the existence of possible worlds in which space and time are
continuous and extend to infinity, and in which each spatio-temporal point
is occupied by infinitely many properties. For the modal realist, this means
that there is an infinity of worlds that apparently exist, but whose existence
cannot be admitted by the linguistic ersatzist. This is a serious theoretical
limitation.
There are only two ways out for the linguistic ersatzist. The first denies
that the possibilities that cannot be described by means of a set of
sentences are genuine, no matter how intuitive they may seem. According
to this response, it would simply be impossible that space is continuous,
because no set of sentences can describe such a scenario. This position,
however, seems ultimately ad hoc and therefore unsatisfactory. The second
option is more promising, though probably still unsatisfactory. According to
it, the requisites of completeness and consistency should be understood
against the background of some ideal agent who possesses a language
much more powerful than the ones we in fact possess; for such an ideal
agent, a possible world where space is continuous would be describable by
means of a set of sentences. Thus, unfortunately we human beings do not
have the ability to express in our languages all of the possible scenarios that
we have reason to believe exist; but some ideal agent would be able to do
so. (For more on this point, see Divers (2002) and Divers and Melia (2002,
2003)).
While the linguistic ersatzist’s solution to the MPP is unimpressive,
she will be able to provide a more persuasive treatment of the EPP. This is
because linguistic ersatzism delivers a surprising result. Normally, knowing
that a sentence is true is importantly distinct from knowing its truth condi-
tions. (Knowing its truth conditions is just what it is to know its meaning,
which in turn is part of what it is to know the relevant language.) Yet, it is
part and parcel of linguistic ersatzism that understanding a sentence about
a possibility is enough for understanding that that sentence is true. So, just
in knowing a language—and therefore knowing the truth conditions of its
sentences—one knows the truth of its sentences about what is possible.
Hence, all it takes to show that we know of the existence of complete and
consistent sets of sentences is, indeed, a theory of what it is to know a
language (including a formal language). Coming to know the fact that Foffo
could have milk and cookies—sentence (5)—amounts to understanding the
conditions through which (5), when expressed in a certain language, is true.
These truth conditions do not require knowledge of the details of an entire
world. That is, in order to know that (5) is true, it is not necessary to know
the details of the complete and consistent sentence that expresses the

9781472524263_txt_print.indd 121 01/12/2015 09:44

 122 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

existence of one of the worlds in which (5) is true (presumably, (5) is true
in many worlds). It would suffice to understand the meaning of the terms
in (5), together with the WLE theory, according to which the truth of (5) is
not possible on its own, but rather through a description of a complete and
consistent set of scenarios.
Before moving on to combinatorialism, let us distill the solution to the PP
proposed by the linguistic ersatzist:

LE: The sentence in natural language: “It is possible that: s” expresses

a maximal and consistent description, given in a certain language,
according to which: s.

§5.3. Combinatorialism
According to the second ersatzist approach we will consider, possible
worlds are not linguistic entities, but are made of the components of our
very world. The approach goes under the label of “combinatorialism”, and is
often associated with the name of David Armstrong, the philosopher who
most prominently and exhaustively developed it in his work of 1989, 1997,
and 2004. Combinatorialism owes its name to the fact that possible worlds
are construed as recombinations of pieces of the actual world: they are thus
worlds much like our own. The recombination schema, as we shall see, in
part follows the principle of recombination offered by modal realism and
discussed in Chapter 4, while also containing some important differences.
Because worlds are ways in which elements of the actual world could be
recombined, we can claim that for combinatorialism worlds are surrogates of
the actual world and, thus, that combinatorialism is an ersatzist theory.
In our analysis of combinatorialism, we shall especially concentrate on
how the theory purports to reply to the MPP. This is because, at the semantic
level, combinatorialism can be squared with most versions of QML, with the
identity of individuals across worlds. In this sense, no particular discussion
of the combinatorialist’s approach to the semantics of modal sentences is
required. Nonetheless, there is a semantic aspect that will turn out to be
relevant during our consideration of the combinatorialist metaphysics: this is
the combinatorialist’s understanding of the relationship of “truth in a world,”
which relates a sentence and a world. We shall come back to this in due
course, towards the end of our presentation of the position.
Combinatorialism explains possibility in terms of actual entities plus the
principle of recombination. The latter is regarded as a metaphysical principle,
much like modal realism. Thus, the combinatorialist’s answer to the EPP will

9781472524263_txt_print.indd 122 01/12/2015 09:44

 ERSATZISM 123

depend on how we come to secure knowledge of both actual entities and,

most distinctly, the principle of recombination. With respect to the latter,
the most plausible strategy open to the combinatorialist follows the one
suggested by Lewis and discussed in Chapter 4: the principle of recombi-
nation is known through the philosophical analysis of the key notions involved
in the principle. It is indeed to such analysis that we shall now turn, while
addressing the combinatorialist answer to the MPP.

§5.3.1. States of affairs

At the metaphysical level, combinatorialism crucially relies on states of affairs.
While “states of affairs” has a generic connotation in English, and may be
regarded as synonymous with the generic “scenario” we have employed
so far, in this context the expression connotes a specific type of entity. To
introduce this type of entity, let us consider the following sentence:

6 Foffo has milk and cookies and Fufi eats fish and potatoes.

According to the combinatorialist, (6) expresses the existence of a certain

state of affairs. This is a complex whole made of various constituents: Foffo,
Fufi, milk, cookies, fish, and potatoes. The constituents are arranged in a
specific order and, for this reason, we can claim that the state of affairs
described by (6) is a structured whole. Not all states of affairs need to be
structured wholes; consider, for instance, the state of affairs expressed by
the sentence:

7 It rains.

Aside from the special case of sentences such as (7), all states of affairs have
some structure, that is, they will have at least two constituents, related in a
specific order and fulfilling specific roles.
For Armstrong, constituents divide into two fundamental categories:
individuals (or “particulars”, but we will stick with “individuals”, as has
been our practice so far in this volume) and universals. Foffo and Fufi are
individuals. Being a cat is a universal. There are different ways of drawing
the distinction between individuals and universals. According to one that
is often employed, the key difference between individuals and universals is
that a universal can exist in its entirety in more than one spatio-temporal
region, while an individual cannot; thus, Being a cat is, in its entirety, simul-
taneously where Foffo is and where Fufi is, but neither Foffo nor Fufi can be
in two places at the same time. Another difference is that only universals
can be instantiated (or: exemplified); thus, individuals instantiate universals,

9781472524263_txt_print.indd 123 01/12/2015 09:44

 124 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

while universals cannot instantiate individuals; nor can individuals instantiate

individuals. For example, Foffo instantiates Being a cat, so does Fufi, but
Foffo and Fufi are not instantiated. Needless to say, many find the distinction
between individuals and universals problematic; many, moreover, find the
very idea of universals problematic; ruling out one or the other notion implies
rejecting combinatorialism, as proposed by Armstrong. For the sake of the
argument, we shall set aside any such objections, granting Armstrong suffi-
cient metaphysical ground to display his metaphysical theory of possibility
(and, hence, his answer to the MPP).
According to Armstrong, once we are equipped with a stock of individuals
and universals, we can provide a full-fledged metaphysical analysis of worlds.
The actual world is a vast array of individuals that instantiate a large number
of universals. Combinatorialism is the view that the individuals and universals of
the actual world can be rearranged. Let us offer an example. It seems possible
that, instead of (6), the following scenario could have occurred:

8 Foffo eats fish and potatoes and Fufi has milk and cookies.

For the combinatorialist, (8) is possible insofar as it is a legitimate recombi-

nation of constituents of the actual state of affairs described in (6). Thus, from
the fact that (8) is possible we can infer that the following sentence is true:

9 It is possible that: Foffo eats fish and potatoes and Fufi has milk and
cookies.

The truth conditions for (9) are explained by the combinatorialist in terms of
the truth conditions for (6). We shall come back to this remark in a moment.
For now, we shall sum up the combinatorialist solution to the MPP as follows:

C: A state of affairs is possible if and only if it is obtained by means of a

legitimate recombination of constituents of at least one actual state of
affairs.

In order to further understand the combinatorialist stance, C must be debated

from different angles. We will study three of them: the notion, invoked by C,
of a constituent; the principle of recombination that sustains C; and the extent
to which non-actual worlds may depart from the actual one.

§5.3.2. Constituents
A recombination is defined based on the constituents of some state of affairs.
As we have seen, according to the combinatorialist, (almost) every state of

9781472524263_txt_print.indd 124 01/12/2015 09:44

 ERSATZISM 125

affairs possesses a metaphysical structure; this is typically established by

appeal to the logical structure of the sentence that expresses the state of
affairs. The logical analysis of the structure of a sentence can be made in
at least two ways: through the identification of logically relevant roles, or
through the identification of logically relevant constituents.
The first modality of analysis involves identifying the key roles of the
sentence that expresses the state of affairs: for example, subject, predicate,
and adverb. Nevertheless, in some cases, this type of analysis does not dig
deep enough into the structure of the sentence: the role of the subject could
be fulfilled by several different individuals; the predicate or adverb could be
compound. From a combinatorialist point of view, this fact is problematic
in that it does not allow us to identify the metaphysical structure of the
sentence, and thus the ultimate elements subject to possible recombination:
for example, it does not allow us to identify the various elements that form a
subject, but rather it forces us to recombine the entire subject.
For this reason, the combinatorialists support a logical analysis based on
the concept of a logical constituent of a sentence. A logical constituent of a
sentence is, intuitively, a term that denotes an object or an action (including
states, such as a frame of mind), or a modification of the object, action, or
state of affairs. Now, an important distinction is the one between atomic and
non-atomic constituents. An atomic individual is such that it has no proper parts,
while an atomic universal is such that it is not a conjunction, a disjunction, or
a structure made out of other universals. For example, in (6) we may suppose
that the atomic constituents are: the predicate “to eat”; the terms “Foffo”,
“Fufi”, “milk”, “cookies”, “fish”, and “potatoes”. On the other hand, in the state
of affairs expressed by (6) there are also non-atomic constituents: “Foffo eats
fish and potatoes” and “Fufi has milk and cookies”. These constituents are not
atomic insofar as they are obtained based on other constituents.
In the simplest model proposed by Armstrong, which shares important
elements with the metaphysical picture provided by Wittgenstein, all the
constituents are atomic. (See Wittgenstein (1921), Skyrms (1981), and
Armstrong (1997).) Now, imagine having a stock of atomic individuals and a
stock of universals such that: (i) each individual in the stock instantiates some
universal and, furthermore, (ii) each universal in the stock is instantiated by
some individuals. Imagine that the world fulfilling these conditions is the
actual world, so that any recombination of constituents of the actual world
delivers a world made of atomic constituents and the totality of recombina-
tions associated with the actual world recombines the totality of individuals
and universals we have in stock. If this is the case, then the actual world is
what Armstrong calls a “Wittgenstein world”.
As long as combinatorialism assumes that the actual world is a Wittgenstein
world, the combinatorial machinery for combinatorialism is relatively

9781472524263_txt_print.indd 125 01/12/2015 09:44

 126 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

straightforward. Unfortunately, this assumption is implausible. For all we know,

some important constituents of the actual world (e.g. physical, chemical, or
biological kinds) may not be atomic, but rather exhibit a variety of characteristic
structures. How are we then to recombine structured constituents? For instance,
should their parts be recombined as well? Consider how we might recombine
a molecule of water: should its atoms of hydrogen and oxygen be recombined
as well, or does only the whole molecule figure in the recombination? Sider
(2005) raises a number of questions along these lines, which still await proper
discussion. Armstrong (2004) argues that combinatorialism may indeed sit best
with atomic worlds, but it is far from an established (scientific or philosophical)
fact that our world is atomic. To see further into this matter, we shall now turn
to the combinatorialist’s formulation of the principle of recombination.

§5.3.3. Recombinations
Armstrong contends that combinatorialism is capable of providing a reductive
analysis of modality on a par with modal realism. The reason is that combina-
torialism explains the meaning of modal sentences in terms of the existence
of non-modal, actual entities. For instance, the truth of (9) is explained in
terms of the truth of (6). Constituents of the actual world are sufficient
to account for all the truths that a theory of modality must explain, or so
contends Armstrong.
Principle C is exempt from some of the difficulties of the concept of recom-
bination observed during our discussion of modal realism, where counterpart
theory forced Lewis to introduce the notion of an intrinsic duplicate. Because
the combinatorialist admits that there is identity among individuals belonging
to distinct worlds (after all, the individuals of other worlds exist as surro-
gates of individuals of the actual world), what she recombines are precisely
individuals of the actual world. On the other hand, the combinatorialist will
face some new challenges, to which we now turn.
Having clarified the notion of a constituent, we can define the concept of
recombination as follows:

R: A state of affairs S is a recombination of certain states of affairs S*

if and only if: (i) all constituents of S belong to at least one of the
states of affairs in S* or are obtained by interpolation or extrapolation
from constituents of at least one of the states of affairs in S*; (ii) S is
distinct from all the states in S*.

A state of affairs recombines other states of affairs if and only if it possesses

a different order of constituents or if it relates constituents that are not related

9781472524263_txt_print.indd 126 01/12/2015 09:44

 ERSATZISM 127

in the original state. The appeal to interpolation and extrapolation allows some
departure from the actual world. A universal is obtained from interpolation of
another when some of its aspects are changed: for instance, imagine that
the gravitational constant had some slightly different value than in fact it has.
A universal is obtained by extrapolation from another one by removing some
of its aspects: for instance, imagine having an electron with no charge. Part
(ii) of the definition is used to guarantee that S is not actual, but it is a mere
possibility. As an illustration of principle R, consider the state of affairs S:

10 Giovanni loves Laura,

and suppose that S is a recombination of the two following states of affairs:

11 Laura loves Pietro,

12 Giovanni admires Gina.

Yet, things aren’t quite that simple. In fact, not all recombinations are
acceptable. For example:

13 Giovanni loves admire

is a recombination of (11) and (12), but it is not acceptable. For this very
reason, C requires that recombinations be legitimate.
In order to explain which combinations are legitimate, it is necessary to
reintroduce the logical analysis based on roles introduced earlier: (13) would
not count as legitimate insofar as “admire” fulfills the role of a term, while
it should fulfill the role of a predicate. Combinatorialism, therefore, is based
essentially on a double logical analysis of sentences, which yields a double
metaphysical analysis of states of affairs: that of atomic constituents and that
of atomic roles.
Thus, contra Armstrong’s contention that combinatorialism offers a
reductive analysis of modality, the appeal to legitimate recombinations makes
C a circular definition of the concept of possibility: possible states of affairs
are those that are possible within the limits of the assigned metaphysical
roles. Surely, this is a theoretical limitation with regard to modal realism; yet,
if we consider that modal realism is the only theory that claims to be able to
define the concept of possibility in a non-circular manner, the concerns raised
by the limitation are minor. The combinatorialist offers us an explanation of
the concept of possibility, but not a definition. And this may be considered
sufficient: in order to offer adequate solutions to the PP, EPP, and MPP, it is not
necessary to eliminate the concept of possibility; it need only be explained.
We can now proceed to formulate the principle of recombination for
combinatorialism:

9781472524263_txt_print.indd 127 01/12/2015 09:44

 128 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

PRC: Any collection of individuals x1, …, xn and universals U1, …, Un could

coexist, or fail to coexist, with any collection of individuals y1, …, yn
and universals V1, …, Vn, as long as: (i) the individuals fit the maximal
external relation of a world, and (ii) the combination delivers states of
affairs.

There are two key differences between the principle of recombination, PR,
seen in Chapter 4 and PRC. To begin with, PRC recombines not only individuals,
but also universals; in fact, PR followed a nominalist metaphysics, where only
individuals exist, while PRC rests on a so-called realist metaphysics, where
universals—as well as individuals—are real. In general, PRC purports to
recombine all constituents of states of affairs (and whether it succeeds in
doing so is debated by its detractors, as we shall see). The second difference
between PRC and PR is PRC’s requirement that the recombination deliver a
state of affairs.
PRC offers us the opportunity to consider the combinatorialist solution
to the EPP. Combinatorialism seems to play upon two distinct types of
knowledge. On one hand, we have the principle of recombination, which
states that all legitimate recombinations are possible. It is a metaphysical
principle, knowledge of which is gained (to a certain extent) independently of
experience. On the other hand, we have empirical knowledge of the atomic
constituents of reality: it is based on this knowledge that we can grasp the
meaning of any modal sentence.
At this time an objection could be made: identifying the atomic constit-
uents of reality is not at all banal. Can a cell be considered an atomic
constituent, or is it composed of other constituents? Are colors, sounds, or
tastes constituents of reality or mere figments of our imaginations? Are gaps
and omissions constituents of reality?
Of course, this objection involves general philosophical problems that are
not exclusive to the combinatorialist, though they do have a greater weight
for her than for the modal realist, for example. The combinatorialist typically
believes that it is not (only) up to philosophers to give an answer to questions
of this type. Combinatorialism brings a conceptual clarity to the PP, but the
fact that it cannot have the last word about what things are possible is not
strange; to find out what is possible, we must also consult other disciplines.

§5.3.4. Alien possibilities

The third aspect of C that requires attention is also related to ways in which
the actual world might depart from a Wittgenstein world. In particular, it
concerns the possibility that the actual world may have contained either fewer

9781472524263_txt_print.indd 128 01/12/2015 09:44

 ERSATZISM 129

or more individuals, or more universals. The simplest circumstance is the

actual world containing fewer individuals; this scenario is typically referred to
as a contraction. The combinatorialist will represent a contraction by means
of a world made of a conjunction of states of affairs that jointly contain fewer
individuals than the ones existing in the actual world.
More difficult is the case of an expansion, that is, a world that contains
more individuals than the actual world. Suppose the stock of recombinable
individuals from the actual world includes n individuals, and we are given the
sentence:

14 There could have been n+1 cats.

Regardless of whether it would turn out to be true or false, (14) seems to

express a state of affairs whose content is straightforward. So the combina-
torialist can hardly appeal to the fact that, despite our modal intuitions, (14)
expresses an impossible state of affairs. Yet how can her theory represent
such a state of affairs if there are not enough individuals in the actual world
to deliver it?
A solution may be to add a clause endorsing the legitimacy of recombi-
nations obtained through the iteration of elements of the actual world: if in
the actual world there is only one pope, a state of affairs with two popes is
possible insofar as it is derived from the actual world by the iteration of a
certain constituent. Thus, a world in which there are more individuals than
there are in the actual world becomes a possibility by iterating some items
of the actual world. Whether iteration should be regarded as a legitimate
operation, from a metaphysical point of view, may nonetheless be doubted.
Despite the potential merits of her view, there is a type of possibility that
the combinatorialist does not seem to be able to explain: alien universals
(which we encountered in Chapter 4). Not only can we think that there
are worlds with more individuals than there are in our own; there could
also be worlds with individuals whose qualities and abilities are completely
distinct from those of the individuals of our world. At one time, such possi-
bilities were excluded by invoking the so-called principle of plenitude for the
actual world, according to which God created (in the actual world) everything
God could have created, not only with respect to number (which is debatable,
but less relevant), but also, and more importantly, with respect to variety.
The fact remains that this principle does not seem to have a clear empirical
confirmation: there are undoubtedly many species of living organism, but why
should we believe that there couldn’t be, or couldn’t have been, even more?
Or why couldn’t the periodic table, for example, have included more elements
than the ones it already includes? In other words: why believe that we live in
the most variegated world that could exist? A combinatorialist cannot account

9781472524263_txt_print.indd 129 01/12/2015 09:44

 130 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

for the plausibility of these questions and yet it is not easy to set them aside.
We are, therefore, left with a serious theoretical limitation.
A final objection to combinatorialism arises from the semantic tie “being
true in”, which relates a sentence and a world. For the combinatorialist,
worlds make-true the sentences that they make true out of necessity; for
instance, (9) is necessarily made true by any world containing the states of
affairs depicted in (9). However, the necessity of the truthmaking relationship
cannot be analyzed in modal terms. Thus, we are stuck with a primitive kind
of modality, despite Armstrong’s contention that combinatorialism manages
to reduce modalities.
To this objection, the combinatorialist may point out that the necessity of
the truthmaking relationship is expressed through a kind of advanced modal-
izing (see Chapter 4), which is not subject to the same sort of analysis as
modal sentences in natural language. It is therefore to be expected that the
theory cannot explain truthmaking in terms of recombinations of actual states
of affairs.

§5.4. Pictorial ersatzism

The two ersatzist positions we have seen up to this point have taken advantage
of the representative power of language and of the notion of recombination
of actual states of affairs, respectively. The third position we will consider—
pictorial ersatzism—is in some respects a cognate theory of combinatorialism
because, ultimately, both theories employ non-linguistic and non-material
structured entities. On the other hand, pictorial ersatzism and combinatori-
alism differ with respect to how the existence of a world is motivated: the
combinatorialist relies upon a principle of recombination rooted in actual
individuals, properties (universals), and laws; but, for the pictorial ersatzist the
maximality and consistency of a world will typically be defined independently
of the laws of the actual world.
The label pictorial ersatzism comes from Lewis (1986: Chapter 3). Versions
of pictorial ersatzism can be attributed already to classic authors, such as
Wittgenstein (1921) and Carnap (1947), and perhaps even Leibniz. After the
development of QML and possible-worlds semantics, pictorial ersatzism
became a predominant position and was developed in different directions by,
among others, Alvin Plantinga (1974a, 2003), Robert Adams (1974), Robert
Stalnaker (1984, 2003), and Peter van Inwagen (1980, 1986). Because of the
large number of authors who have endorsed pictorial ersatzism, there is no
uncontroversial taxonomy of its variants. Our discussion will distinguish three
of the main versions: according to the first, worlds are maximally consistent

9781472524263_txt_print.indd 130 01/12/2015 09:44

 ERSATZISM 131

sets of propositions; according to the second, worlds are complex states of

affairs (but not recombinations of actual states of affairs); according to the
third variant, possible worlds are images. Let us analyze the variants in this
order.

§5.4.1. Worlds as sets of propositions

The first variant of pictorial ersatzism proposes one of the most intuitive
views of possible worlds, according to which they are sets of propositions.
Let us begin by reasserting the basic distinction between sentences and
propositions. Consider the following two sentences of, respectively, English
and Italian:

15 It rains.

16 Piove.

Now, (15) and (16) are clearly not the same sentence—they even contain
a different amount of words; on the other hand, they convey the same
content—that it rains. We call the content of a sentence the proposition
expressed by the sentence.
Insofar as we have reason to believe that our sentences and our thoughts
do convey some content, we have reason to believe that they express propo-
sitions. Or, so claim the authors who endorse the existence of propositions.
Since we seem to have good reasons to believe that propositions exist—
reasons that are independent of our theorizing about modality—the proposal
to analyze modal sentences in terms of propositions is far less controversial,
from a metaphysical point of view, than most other competing proposals on
the market, including modal realism. Because of this, some authors (e.g.
Divers (2002) and Melia (2003)) included this variant of pictorial ersatzism
among the so-called quiet moderate realist positions, in contrast to the
extreme realism of David Lewis.
The defender of the first variant of pictorial ersatzism takes the theory of
propositions one step further. She contends that the content of our modal
sentences can be explained by fine-tuning our theory of propositions. This
operation requires several adjustments and different authors have suggested
alternative ones (see Adams (1974) and Stalnaker (1984) for some examples).
In our view, it takes at least four adjustments to a basic theory of propositions
in order to produce a suitable pictorial ersatzist account of possible worlds. (i)
We must make sure that the propositions out of which worlds are constructed
are neither incomplete nor contradictory (as would be the content expressed
by sentences such as, respectively, “Socrates thinks about” and “Socrates

9781472524263_txt_print.indd 131 01/12/2015 09:44

 132 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

is and is not Athenian”). Some deny that propositions of this sort exist tout
court, but we need not be so demanding, as it suffices to require that no
such proposition is part of a world. (ii) We must require that propositions exist
independently of the existence of actual entities and that they exist indepen-
dently of a language expressing them. This adjustment allows us to claim that
a proposition about non-actual entities (e.g. unicorns) is true or false even if
there actually are no such entities. This adjustment allows modal sentences
that express mere possibilities, including alien possibilities, to be meaningful.
Moreover, it grants the pictorial ersatzist with a basis for maintaining that there
may be some scenarios that are so alien to us that we cannot express them
in our language. (iii) We shall define a world as a set of propositions, since one
of them will typically not be enough to fully capture all that there is to a world.
Such a set should be maximal and consistent, as also required by the linguistic
ersatzist (and a similar requirement will hold for the other two variants of
pictorial ersatzism). And, it is important to remark that these requirements
render the analysis of modality circular. (iv) We shall define a relationship of
“truth at” as holding between sentences and (maximal and consistent) sets
of (complete and non-contradictory) propositions. To illustrate, the sentence:

17 It is possible that: Foffo plays with Fufi

is true if and only if there is a maximal and consistent set of propositions

such that, at such a set, (17) is true; and (17) is true at that set when (17) is
included in the set.
The key tenet of the first variant of pictorial ersatzism can be summed up
as follows:

PEP: The sentence in natural language: “It is possible that: s” expresses

a proposition p, and is true if and only if there is a maximal and
consistent set of (complete and non-contradictory) propositions at
which p is true.

The view is pictorial because possible scenarios are specific entities (i.e.
specific sets of propositions) that depict how the actual world could have
been. The theory employs three metaphysical primitives: propositions, sets,
and the relation “true at”. In a way, these primitives are common to many
solutions of the MPP, regardless of the details of any theory of modality
providing such solutions. On the other hand, pictorial ersatzism is far from
innocent, for two chief reasons. First of all, it compels us to buy into a theory
of propositions as entities that exist independently of actual entities. Second,
if the theory is to have some plausibility, it must endorse a view of proposi-
tions as structured entities of some sort, so that the truth of a proposition

9781472524263_txt_print.indd 132 01/12/2015 09:44

 ERSATZISM 133

is explained by recurring to its structure. (For example (17), if true, is true in

virtue of the structure of the proposition it expresses—a structure made up
of Foffo, Fufi, and the relation between them.) To many philosophers, these
two features alone are reason enough to look elsewhere for a solution to the
PP and its sub-problems.

§5.4.2. Worlds as sets of states of affairs

The second variant of pictorial ersatzism resembles the first, but its key
metaphysical primitive is not the proposition. A theory resting on proposi-
tions, as we have seen, starts off as intuitive, by insisting on the tie between
sentences and propositions; but it ends up introducing unlikely constraints
on our view of propositions, which depict them as abstruse and perhaps
even obscure entities. Plantinga (1974a) and subsequent authors opt for an
ersatzist position where the role of worlds is played by states of affairs.
We encountered states of affairs during our discussion of combinato-
rialism. While the latter was developed after, and under the influence of,
the pictorial ersatzist variant under consideration, the views are distinct.
The two chief differences between them regard the relationship between
actual entities and the constituents of states of affairs, and the nature of the
constituents themselves. Let us discuss them in order.
While combinatorialism crucially relied upon a principle of recombination
that operated upon actual individuals and universals, no such appeal is
required in the variant of pictorial ersatzism under consideration. That is,
combinatorialism delivers only non-actual states of affairs whose constituents
are (obtained by iteration, interpolation or extrapolation from) constituents
of actual states of affairs; the variant under consideration, instead, allows
for constituents that are non-actual. Thus, the variant cannot regard states
of affairs as endowed with causal powers or at least as part and parcel of
the causal structure of reality; on the other hand, it can accommodate the
existence of alien possibilities more readily than combinatorialism.
Second, the variant under consideration explains the existence of non-actual
constituents by postulating that constituents of states of affairs are necessary
existents, endowed with individual essences. Suppose that Fufi, the cat,
could have had a twin sibling, Fefe. Fefe is an unactualized cat. According to
the variant under consideration, it is necessary that there is a state of affairs
of which Fefe is a constituent and, in this sense, Fefe exists out of necessity.
Moreover, Fefe has an individual essence, that is, there are certain properties
of Fefe that define what sort of entity it is (among other things, a cat and a
sibling of Fufi): Fefe retains those properties in any state of affairs and it is
identified by means of them (for a discussion of essentialism, see Chapter 3).

9781472524263_txt_print.indd 133 01/12/2015 09:44

 134 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Equipped with the theory of states of affairs just outlined, we can define
worlds as maximal and consistent sets of states of affairs. More explicitly:

PES: The sentence in natural language: “It is possible that: s” expresses

a state of affairs S, and is true if and only if there is a maximal and
consistent set of states of affairs at which S is true.

The analysis of modalities provided by PES is even more circular than the one
proposed by PEP: it relies not only on maximality and consistency, but also on
individual essences and necessary existent individuals. We shall come back
to necessary existents in Chapter 8, both in conjunction with Meinongianism
(the view that there are certain things that do not exist) and when considering
the theory of necessary existents suggested by Timothy Williamson.

§5.4.3. Worlds as images

The third pictorial ersatzist proposal envisions worlds as images, as opposed
to representations that possess a logical or linguistic structure. Images, here,
do not stand for figures embodied in some material object or drawn by some
agent, but for figures that exist independently of such material inscriptions.
After all, it is still possible that a prism with 30,827 sides could exist, even
if—say—no material object has the form of such prism and no one has ever
drawn it. Why, then, should we not believe that there is an image for every
possible world and that possible worlds are nothing but those images, no
matter whether the images represent some actual material inscription?
The chief advantage of pictorial ersatzism over the other ersatzist positions
we have seen lies in the fact that pictures have a greater expressive power
than descriptions or recombinations. By embracing pictorial ersatzism we can
assert that there are worlds with a cardinality of individuals similar to that
of the continuum, just as we are able to represent worlds with a variety of
individuals that are not attainable through the recombination of actual varieties.
That said, pictorial ersatzism must admit that it provides, not a definition
of possibility, but, more modestly, an explanation of it. The figures that
represent possible worlds are those whose interpretation is consistent and
complete. A figure, in and of itself, does not constitute a representation;
the representation is obtained only once the figure has been interpreted.
(Similarly, a figure, in and of itself, is not a map; we have a map only once the
criteria for interpretation of the figure have been provided.) But consistency
and completeness (as seen in the foregoing discussion of linguistic ersatzism)
are modal concepts: that is, they are concepts that presuppose an under-
standing of possibility.

9781472524263_txt_print.indd 134 01/12/2015 09:44

 ERSATZISM 135

The notion of interpretation raises a second problem. From what has been
said, it emerges that a possible world cannot consist simply of a figure, but
of an interpreted figure. Thus, pictorial ersatzism’s solution to the PP would
be the following:

PEI: The sentence in natural language: “It is possible that: s” expresses

an interpreted figure such that: the figure represents s; and, according
to the interpretation, the figure constitutes a maximal and consistent
representation.

The problem is that PEI itself lacks a clear interpretation. What are interpreted
figures? In what sense can we consider them to be surrogates for entities of
the actual world? Moreover, doubts could be raised with regard to how the
figures to which PEI refers are known. Certainly, these questions will have
to be adequately answered if this version of ersatzism is to attract broader
consensus.

§5.5. Atomic ersatzism

The final ersatzist position that merits a closer look is atomic ersatzism. This
can seem to be a strange position at first, because it rests on a very simplistic
metaphysical proposal, according to which possible worlds are atomic, simple
entities. That is, according to atomic ersatzism, the answer to the MPP is
straightforward and can be summed up as follows:

MAE: A scenario is possible if and only if there is an atomic entity that is

abstract and maximal and that represents that scenario.

According to the atomic ersatzist, worlds have no structure at all; they seem
to represent possible scenarios by magic: hence the label that Lewis reserved
for this position: magical ersatzism.
Up until this point in our discussion of the ersatzist positions, we have
spoken of worlds as entities that have certain constituents: elements of a
language, abstract correlates of concrete constituents of the actual world,
propositions, states of affairs, or images. But we might imagine that a world
is a single entity, without real parts. Of course, we can imagine an object
being divided into parts, but this does not mean that the object is dividable:
for example, we can think of a quark being divided into parts, but this does
not mean that it can be done. In other words: metaphysical atoms do not
necessarily coincide with topological or conceptual atoms.

9781472524263_txt_print.indd 135 01/12/2015 09:44

 136 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

For the atomic ersatzist, as in the case of any other ersatzist position,
actuality is metaphysically privileged. There are many worlds. Each of them is
a maximal entity in the sense we defined (for the other ersatzist positions).
Only one of them is actual. All of the others are abstract. But they are not
obtained through a recombination of parts of the actual world, parts of
figures, or parts of a language. They exist; period. And they represent the way
in which our world could have been because they are distinct from our world.
Their being distinct is explained in terms of their representational powers,
which vary from world to world, but which cannot be analyzed in terms of
constituents.
The appeal to representational power raises some perplexities. Given that
worlds are metaphysical atoms, we cannot say much about them. We know
that they are atomic: that is, they do not have parts. We know that they are
maximal: that is, it is not possible to add any part to any world. We know that
they are abstract. But, what are they abstractions of? It appears that, as we
saw with pictorial ersatzism, atomic ersatzism also requires that its preferred
metaphysical material be interpreted in order to deliver possibilities. MAE
must, therefore, be reformulated as follows:

MAE*: A scenario S is possible if and only if there is an atomic entity

S* that is abstract and maximal, and of which there is at least one
interpretation according to which S* counts as a representation of S.

MAE* contains a double appeal to modal concepts: the maximality and the
validity of representation. Furthermore, given that S* is atomic, there is no
reason that it should count as a representation of S: it represents it because
it represents it. By virtue of this distinctive feature of the theory, the repre-
sentational powers of each world seem magical.
The solution to the PP proposed by atomic ersatzism can now be stated:

AE: The sentence in natural language: “It is possible that: s” expresses

an atomic entity S* that is abstract and maximal, and of which there
is at least one interpretation according to which S* counts as a
representation of the scenario expressed by the sentence “s”.

Some words are needed in order to explain the plausibility of atomic

ersatzism. In this chapter, we have examined five ersatzist conceptions of
possible worlds, in addition to atomic ersatzism. Each of them suggested
complex analyses of the elements of which a world is allegedly composed.
And yet, we have seen that thinking of such elements as linguistic entities,
images, structured propositions, or structured states of affairs leads to compli-
cations regarding the kinds of entities worlds are or regarding the extent of

9781472524263_txt_print.indd 136 01/12/2015 09:44

 ERSATZISM 137

the possibilities that worlds are able to represent (including especially alien
possibilities). The atomic ersatzist takes a different route. She maintains that
worlds are simple entities. Thus, the issue of alien possibilities is solved: we
can have as many simple entities as we need in order to represent all the
possibilities that the theory should. Explaining the representational powers of
a world is, of course, the most delicate aspect of atomic ersatzism. But, just
as a single symbol, such as an ideogram, can stand for many words and, as
such, for some complex scenarios, so a single abstract entity can stand for
the complex scenario that is a possible world.
While the other ersatzists have tried to explain the complex scenario that
is a world by complicating the metaphysical structure of worlds themselves,
the atomic ersatzist discharges that responsibility by complicating the relation
of representation between a world and a scenario. Yet such representation
happens by magic and cannot be explained in metaphysical terms, as we
have seen; and for this reason atomic ersatzism may be regarded as implau-
sible. On the other hand, atomic ersatzism can describe the representation
by means of semantics: the representation consists in all the sentences
that a certain world makes true. Furthermore, it may by now be clear that
the metaphysics of modality forces any account to adopt some-or-other
unwelcome (at times incredible) tenet; atomic ersatzism is, thus, in good
company in making a seemingly bizarre metaphysical choice, and we should
at least concede that atomic ersatzism cleverly exploits a theoretical possi-
bility that had not yet been seen among the theories in this volume.

Study Questions
• What is the meaning of the term “ersatz”? And why is it appropriate for
characterizing ersatzism?
• What is the main characteristic of ersatzism at the metaphysical level?
• What is the main contention of linguistic ersatzism?
• How does linguistic ersatzism generate possible worlds?
• What are the chief drawbacks of linguistic ersatzism?
• What, according to a combinatorialist, is a state of affairs?
• What, according to a combinatorialist, is a constituent?
• What is a Wittgenstein world?
• What is the principle of recombination in combinatorialism?
• What are the principal difficulties for combinatorialism?
• What is a proposition?
• What are the main difficulties for using propositions in conjunction with
pictorial ersatzism?
▲

9781472524263_txt_print.indd 137 01/12/2015 09:44

 138 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

• What is the distinctive trait of states of affairs, for the pictorial ersatzist
variant that employs them?
• What is the advantage of supposing that worlds are images?
• What are the drawbacks of supposing that worlds are images?
• How does atomic ersatzism explain the meaning of a modal sentence?
• What motivates the supposition that worlds are atomic?

FURTHER READING
Along with modal realism, ersatzism is the most debated position in the philosophy
of modality. There is a crucial difference between the two theories, though: while
modal realism has been openly endorsed by only a handful of philosophers, plenty
are those who have professed allegiance to ersatzism or advanced a theory that
seems ersatzist. It is for this reason not uncontroversial to systematize the spectrum
of ersatzist positions, and the reader is invited to explore the debate along several
paths. For an introductory survey of the various ersatzist positions, see Lewis
(1986: Chapter 3), Divers (2002: Part Three), and Melia (2003: Chapters 6 and 7).
Also useful to consult are Sider (2002) and Stalnaker (2012). Some classic readings
include: Carnap (1947), Cresswell (1972), Plantinga (1974a, 2003), the collections of
essays in Linsky (1974), and Loux (1979), Kripke (1980), and Skyrms (1981).
For a discussion of trans-world identity, see: the texts collected in Loux
(1979), Kripke (1980), van Inwagen (1985), and Lewis (1986: Chapter 4).
For a formulation of linguistic ersatzism, see Roy (1995) and Stalnaker (1984,
2004). For a critical discussion, see McMichael (1983), Lewis (1986, §3.2.), and
Heller (1996, 1998).
For a systematic formulation of combinatorialism, see Armstrong (1986,
1989). Important follow-ups are Armstrong (1997, 2004). For a critical assessment
of the position, see Kim (1986), Bradley (1989), Lewis (1992), Thomas (1996), Sider
(2005), Efird and Stoneham (2006), and Wang (2013).
For a formulation of the various pictorial ersatzist positions, see Adams (1974),
the essays contained in Loux (1979), Stalnaker (1984, 2003), van Inwagen (1980,
1986), and Plantinga (1974a, 2003). For a critical discussion, see especially Lewis
(1986, §3.3.). Also useful is Brogaard (2006).
For a critical presentation of atomic ersatzism, see Lewis (1986, §3.4.). For
a defense of the position, see Denby (2006). Also in connection with atomic
ersatzism, it is useful to consult Brogaard (2006).

9781472524263_txt_print.indd 138 01/12/2015 09:44

 6
Modal fictionalism and
modal agnosticism

F ictionalist strategies in metaphysics have been endorsed with respect to a

wide variety of subjects, ranging from meta-ethics to truth and the ontology of
mathematical entities; the first half of this chapter considers how fictionalism has
been applied to modalities. First, modal fictionalism is illustrated and three different
formulations of it are given (§6.1.); then, five problems for modal fictionalism are
discussed (§6.2.). In the second half of the chapter, a more recent modal theory
is taken up—modal agnosticism—which proposes to delimit the range of modal
sentences expressing true or false propositions. After a distinction between
strong and moderate variants of modal agnosticism (§6.3.), two objections to this
view are examined (§6.4.).

*****

I n this chapter we will consider the final two theories of possibility that
are based on possible-worlds semantics: modal fictionalism and modal
agnosticism. Both develop what seems to be the most immediate advantage
of possible worlds: references to them do not have to be taken literally and,
thus, we need not make special room for worlds in our conception of reality.
This feature of possible worlds evidently runs counter to the theoretical core
of both modal realism and ersatzism, which aimed—each in its own way—to
provide a suitable metaphysical and epistemological story to pair up with the
semantic benefits of possible-worlds semantics. Furthermore, both modal
fictionalism and modal agnosticism were formulated after modal realism and
ersatzism: the first article to defend a fictionalist theory of modality is typically

9781472524263_txt_print.indd 139 01/12/2015 09:44

 140 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

considered to be Rosen (1990) (Rosen’s view being further developed in 1993

and 1995), though Armstrong (1989) also discusses the hypothesis of a fiction-
alist theory which, as we will see, differs from Rosen’s theory in several ways.
Agnosticism had not received adequate consideration before Divers (2004).
Let us now take a closer look at the two positions. The first half of this
chapter discusses how fictionalism has been applied to modalities (§6.1.) and
why that’s problematic (§6.2.). In the second half of the chapter, we will take
up modal agnosticism (§6.3.) and point out some difficulties for it (§6.4.).

§6.1. Fictional worlds

In recent years fictionalist approaches to metaphysical questions have
received considerable attention in the analytic philosophical community, and
have been developed in diverse philosophical areas: from the philosophy
of science to the philosophy of mathematics, and from the commonsense
ontology of objects to modality and ethics. (For a general overview, see
Kalderon (2005) and Eklund (2007).) The distinguishing feature of modal
fictionalism is that it considers what is arguably the most useful discourse
we have for explaining modality—possible-worlds semantics—to be a useful
fiction. (For a comparison between modal fictionalism and modal ersatzism,
see Brogaard (2006).)
Before examining the details of the theory itself, we must make a clarifi-
cation. It is rather ambiguous to say that a certain discourse must be treated
as a useful fiction, because it is unclear which aspects of the discourse should
be regarded as fictional. It could be hypothesized, that all fictions are nothing
but discourses introduced for the purpose of explaining other discourses.
When we say that Pinocchio is a piece of literary fiction, we do not believe that
the things The Adventures of Pinocchio tries to shed light on—certain traits of
human nature, let’s say—are also fictional. Thus, Pinocchio is a fiction in the
sense that the sentences contained in the book are not literally true or false;
but certain other sentences, closely connected to those contained in the text,
seem to be true.
Understanding this point is crucial for understanding the modal fictional-
ist’s theory and its difficulties. The modal fictionalist does not claim that modal
sentences are fictional in and of themselves; we are not feigning when we
say that Foffo could have had milk and cookies; we are feigning when we try
to explain that scenario through a language that makes reference to (possible)
worlds. Therefore, the modal fictionalist typically does not deny the truth of
what theories considered to be fictional try to explain, but simply the truth
of those theories.

9781472524263_txt_print.indd 140 01/12/2015 09:44

 MODAL FICTIONALISM AND MODAL AGNOSTICISM 141

With this clarification under our belt, we can now move on to the formu-
lation of the fundamental theory of modal fictionalism, which sets out to
resolve the PP as follows:

F: The sentence in natural language: “It is possible that: s” is true if and

only if its orthodox possible-worlds analysis is true.

For instance,

1 Foffo the cat could have had milk and cookies

is true only if, in the fiction of a theory of possible worlds,

2 There is a world in which Foffo the cat has milk and cookies

is true.
In other words, modal fictionalism tries to join together two observations,
both of which appear to be valid, but which are in tension with one another: (i)
modal sentences need a theory in order to be explained and the best one on
the market seems to be the theory that makes reference to possible worlds;
(ii) theories that take the existence of possible worlds seriously encounter
significant problems when trying to explain the nature of the entities to which
they refer. The modal fictionalist sets out to use a theory that makes reference
to possible worlds in order to explain the meaning of modal sentences, while
at the same time denying that the theory should be taken seriously. Is this
feasible?
Before taking into consideration some of the numerous objections that
have been raised against modal fictionalism, we must make certain clarifica-
tions and elucidations regarding F. Let us begin with the following expression:
“in the fiction of a theory of possible worlds”. The lack of reference to a
specific theory should be noted. In fact, among the theories of possible
worlds illustrated in the previous two chapters, the fictionalist can afford
to choose whichever she prefers. They are all fictions. But how is she to
choose?
After weighing the pros and cons, it seems it would be a good idea to
choose modal realism, as Gideon Rosen did when he first formulated modal
fictionalism, in 1990. In fact, as we have seen, there are few doubts regarding
the greater conceptual potential of modal realism compared to the various
ersatzist proposals: modal realism is able to account for a greater number of
possibilities (see Chapter 4 and the cardinality objection in §5.2). Therefore,
if we were to choose a theory independently of what the theory suggests
exists, modal realism would come out on top—hence, Rosen’s decision. It
is, nonetheless, important to clarify that the fictionalist is akin to a spectator

9781472524263_txt_print.indd 141 01/12/2015 09:44

 142 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

on the debate between theories of possible worlds, and may at any point
change sides. Typically, fictionalists choose whichever theory proves most
conceptually potent.
Another element of F that must be clarified is the reference to an undeter-
mined fiction of possible worlds and, therefore, the translation of s into s*.
Rosen suggests that the theory may be determined using a simple prefix—
“according to the story …”—to be placed in front of an exposition of the
relevant theory—such as, for him, the modal realism of Lewis (1986). The
result is the following principle, in which s stands for a sentence in ordinary
language and s* stands for the orthodox possible-worlds analysis of s:

RF: The sentence in natural language: “It is possible that: s” is true if and
only if, according to the story told in Lewis (1986), “s*” is true.

Rosen’s strategy is clever and points to an anti-realist solution of the PP

that had been overlooked until 1990. Possible-worlds semantics allows the
most sophisticated logical and semantic treatment of sentences embedding
modalities; however—as modal skeptics, modal expressivists, and modalists
have pointed out—possible-worlds semantics raises a host of metaphysical
and epistemic problems that are difficult to solve. An ideal compromise is
to retain the depth of analysis proper to possible-worlds semantics, while
avoiding the metaphysical commitment. That is the aim of modal fictionalism.
The solution is hence found in endorsing the best comprehensive theory
of (possible) worlds—that is the theory that provides the most persuasive
answers to SPP, MPP, and EPP—while regarding that theory as a useful
fiction.
But this is not the only way we can interpret the fictional framework. We
could assume, following a suggestion of Armstrong (1989), that there is more
than one framework: not only is the whole of modal realism a story; it is a
story made of infinite stories: one for every possible world. The exact formu-
lation of the F principle would, therefore, be the following:

AF: The sentence in natural language: “It is possible that: s” is true if and
only if, according to the story told in Lewis (1986), there is a collection
of stories such that, according to at least one of them, s’s equivalent
“s*” is true.

If, on the other hand, s is a sentence that expresses the necessity of a

situation, then s is true if and only if according to all the stories within the
collection (i.e. all of the worlds) s* is true. The idea, then, is that the fiction of
the modal fictionalist has two layers: the first being Lewis (1986), the second
being each world countenanced by modal realism. Two remarks about AF are

9781472524263_txt_print.indd 142 01/12/2015 09:44

 MODAL FICTIONALISM AND MODAL AGNOSTICISM 143

in order. First, it is important to remark that “collection” does not refer to an

entity in and of itself, but should be understood simply as a list of stories.
Second, as we shall see later on, adding a second layer is not free from
important theoretical consequences. On one hand, it provides a clear-cut
fictionalist interpretation of each world within the modal realist framework; on
the other hand, it risks weakening modal fictionalism enough to make it less
attractive than RF. Ultimately, then, the choice between RF and AF is not free
of theoretical consequences and may resolve in favor of the former.
A general point about modal fictionalism that should be highlighted is the
fact that F (and its variants RF and AF) maintains that the truth of a sentence
in a fictional discourse has the role of establishing the truth of a modal
sentence in ordinary language. At first sight, this comes off as a controversial
affirmation: after all, a fiction is not true; nor is it necessarily false; it is simply
a fiction. Therefore, the notion that the truth of our modal sentences is set
by a fiction certainly comes off as suspicious. We could make a criticism
such as this: possible-worlds semantics unquestionably assigns certain truth
values to modal sentences; yet, for fictionalists, possible-world semantics is
nothing but a fiction; if fiction tells us that a certain sentence is true or false,
why should we believe that it in fact is?
Upon more careful reflection, the affirmation contained in F may appear
more plausible. In a certain sense, it depends on our basis of comparison.
Admittedly, we may be unwilling to concede that—absent an adequate inter-
pretation—that which is written in the story of Pinocchio could be directly
translated into truths concerning human nature: after all, the noses of liars
do not grow, and we don’t just happen to bump into the cat, the fox, or the
fairy with turquoise hair. However, our perspective may be different if we
consider fictions that are put to different uses. For instance, take the case
of an interior decorator, who speaks of sofas, chairs, tables, empty spaces,
etc. while furnishing a house; the decorator is also employing a fiction, which
is useful to furnishing a house. However, the decorator’s case seems to
be different from Pinocchio’s: even while referring to fictional entities, the
decorator establishes what we must or mustn’t do in our house. Why, then,
should we not assume that the case of the modal realist is more similar to
that of the decorator than to that of Pinocchio? By simply using a fiction, the
modal realist, like the decorator, determines the truth regarding a certain
discourse—modal discourse.
In conclusion, the distinctive answers to the SPP, MPP, and EPP provided
by the modal fictionalist consist in borrowing the answers of another theory—
typically, modal realism—and treating those as fictions that are useful for the
philosophical purpose of solving the PP. Although such a move may seem
innocuous at first, we shall see that it is not. To establish itself as a plausible
theory of modality, modal fictionalism needs to supply many more details, in

9781472524263_txt_print.indd 143 01/12/2015 09:44

 144 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

particular with respect to the metaphysics and epistemology of the fictions

that allegedly found modal discourse. Whether RF or AF or WF is the principle
of choice will make a big difference already; the detailed interpretation of each
principle will then decide the plausibility of the modal fictionalist proposal.

§6.2. Five problems for modal fictionalism

The modal fictionalist program has been criticized in many respects. We
will highlight five of them here. The list of objections we will consider is not
exhaustive and we will privilege some underrepresented ones. In fact, a
number of objections were considered already by Rosen (1993) and Brock
(1993) themselves; most notably, this includes the so-called Brock-Rosen
objection, which in brief goes as follows. According to the modal fictionalist,
at all worlds it is true that there are many possible worlds. But then—by
the definition of necessity as truth in all possible worlds—, according to the
modal fictionalist, it is necessary that there are many possible worlds. This
is precisely what the modal fictionalist denies, however. It is still debated
whether fictionalism can solve this objection (cfr. Nolan 2011 for an overview)
and the subtleties of the dispute cannot be dealt with here.
The first criticism we will consider claims that the very formulation of
modal fictionalism makes use of modal concepts. The particular part of F that
has been in the limelight is, “in the fiction of a theory of possible worlds”.
Some have found in it an implicit reference to a counterfactual; in fact, the
best way to interpret the passage would be the following: “if the fiction
of a possible-worlds theory were to be realized”. (This criticism is already
considered in Rosen (1990), but see also Divers (1995, 1999).) In other words,
according to the criticism, F would be better interpreted as follows:

F*: The sentence in natural language: “It is possible that: s” is true if and
only if is true if and only if, were the orthodox possible-worlds analysis
to be true, then the sentence would also be true.

In light of what we have previously said, we are able to clarify why the
modal fictionalist would not be willing to accept F*. According to the modal
fictionalist, the critic who proposes F* does not fully appreciate the strength
of F. Modal fictionalists do not believe that the truth of sentences in ordinary
language is determined by the sentences of possible-worlds semantics,
provided said theory were true; rather, the truth of sentences in ordinary
language is determined by what is stated in fiction, without any suppositions.
Therefore, F does not call upon any modalities. It should also be noted that

9781472524263_txt_print.indd 144 01/12/2015 09:44

 MODAL FICTIONALISM AND MODAL AGNOSTICISM 145

the interpretation F* could be accepted by a proponent of a bland version of

modal fictionalism—such as Divers (1999). In fact, for these supporters, modal
fictionalism does not claim to determine the truth of modal sentences in
ordinary language, but simply to explain them; and F* could be an explanation.
A variant of the first criticism is raised by Bob Hale (1995a, 1995b), and
concerns the modal status of the F principle. Though this criticism is also
related to F*, it is often presented independently of the criticism we have
just considered. Hale (1995b) states that the following question presents a
dilemma for the modal fictionalist: is the fiction referred to in the F principle
contingently or necessarily false? If it is necessarily false, then F* is not a
plausible interpretation of F, insofar as the antecedent of the counterfactual (“if
that theory were to be realized”) would be false in any case—and, therefore,
by definition, the consequent would be true. If it is contingently false, then the
sentence “Possibly, the possible worlds fiction is true” is true. If we apply to
such sentence the fictionalist’s analysis suggested in F*, then we obtain the
following: “If the possible worlds fiction is true, then there is a world where the
possible worlds fiction is true.” Now, the latter is trivial, but the claim that the
possible worlds fiction is possibly true is not trivial! Thus, when we suppose
(per the second horn) that the possible worlds fiction is contingently false, F*
gives us the wrong analysis of “Possibly, the possible worlds fiction is true”.
However, Hale’s question does not seem to pose a true dilemma for the
modal fictionalist. As previously clarified, the modal fictionalist would deny
F*: in and of itself, fiction determines the truth of modal sentences in ordinary
language. Whether fiction is contingently or necessarily false is not a question
that affects F in any way, insofar as it does not affect the role that fiction
would have in establishing the truth of modal sentences in ordinary language.
The second criticism, proposed in Peacocke (1999: 154), concerns the
choice of fiction. Of all imaginable fictions, F must choose some theories
that make use of possible-worlds semantics (one of the theories illustrated in
the previous two chapters): but why make that choice in particular? Is there
perhaps an implicit obligation in F to do so?
This criticism naturally suggests a delicate point which has not been
sufficiently stressed in the extant literature. By way of explaining her choice,
the modal fictionalist could appeal to the greater conceptual capacities of
possible-worlds semantics. In and of itself, however, this is insufficient reason
to land on modal realism in particular, as many modal fictionalists do. It seems
that their choice implies an appeal to plausibility, which it is fair to make
explicit:

F**: If we wish to judge in the most apparently philosophically plausible way

the truth of the sentence in natural language: “It is possible that: s” then

9781472524263_txt_print.indd 145 01/12/2015 09:44

 146 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

we must assert that said sentence is true if and only if its equivalent “s*”,
in the fiction of a possible-worlds theory, is true.

But F** contains a modal term, “must”; therefore, the fictionalist is implicitly
relying upon modality when formulating his principal theory: how are we to
interpret this implicit modality?
Alas, there is no compelling answer to this question to date. The modal
fictionalist could bite the bullet and accept that there is a modal assumption
in his theory that cannot in turn be analyzed as a fiction; after all, as we have
seen, some primitive modalities have to be accepted not only by modalists
and ersatzists, but also by modal realists; the modal fictionalist would simply
be joining the bandwagon. Still, if the point highlighted by F** is correct, then
the most appealing trait of modal fictionalism—deferring the analysis of all
modal talk to the best theory of modality on the market—is undermined, and
we are left with little clue as to how the modal fictionalist’s core principle is
supposed to be interpreted.
The third criticism relates to the relevance of the fiction invoked in F and
has so far gone unnoticed in the literature. How are we to believe—one
might object—that a fiction can have such a relevant role in our lives that it
can determine—say—whether Elena should be punished for drunk-driving
because, according to a fiction, the fictional character representing her does
harm someone? More important to us is what Elena does in our world, not
what happens in some fiction, no matter how useful such a fiction may
be claimed to be. Ultimately, modal fictionalism commits the same fallacy
committed by the counterpart theorist (encountered in Chapter 4, when
discussing modal realism): explaining possibility in terms of entities that are
not the ones for whose behavior we sought an explanation. We want to know
how to judge the entities of our world, and we are instructed to consider
other entities—other-worldly ones in the case of counterpart theory, and ficti-
tious ones in the case of modal fictionalism.
The modal fictionalist, like the modal realist, holds the cards in terms of
devising a response to this criticism. Ultimately, F is an explanation of the
meaning of our modal sentences; it would seem legitimate for an explanation
to involve entities we had not foreseen when we raised the question. If we
go to the doctor with a simple stomachache and she, in response, brings
up entities we have never heard of before (certain types of bacteria and
enzymes), which we were unaware had anything to do with stomachaches, it
would not come as a surprise to us. Many accept principles that explain who
we are and how we must act in this world in terms of what we have done
in previous lives, or of what a deity mandates; yet, before learning of these
explanations, we never thought of those previous lives or of the deity and its
teachings.

9781472524263_txt_print.indd 146 01/12/2015 09:44

 MODAL FICTIONALISM AND MODAL AGNOSTICISM 147

We should not be misled by one of the critical features of the dispute over
possibility: the fact that possibility must, inevitably, find solutions that go
far beyond what we can establish by empirical observation, if it is to take a
cognitivist stance (a point we illustrated in Chapter 2, when discussing radical
modal skepticism). If the explanations of the modal realist and modal fiction-
alist do not prove to be convincing insofar as they involve entities that are not
causally related to those of our world, this has more to do with the nature of
the philosophical problem than with the specificity of the theories that stand
accused. Not every theory of possibility can speak of causal explanations, as
the doctor or the theologian does; therefore, from this point of view, modal
realism and modal fictionalism are neither more nor less prone to criticism
than their rival theories. And the fact that they involve entities that are distinct
from those invoked when the problem is raised is not a unique feature of
these theories: doctors and theologians, among others, do the same.
The fourth criticism, raised in Nolan (1997: 264–6) and more recently in Kim
(2005), concerns the ontological status of the fiction invoked in F. We might be
wondering whether it is a fiction in the sense that Pinocchio is fictional or in
the sense applicable to a non-Euclidean geometry. In other words: according
to the modal fictionalist, is the modal realist’s fiction an artifact (such as a
character invented by an author), or is it an abstract and eternal entity, whose
existence is independent of our conceiving it, which the modal realist finds
himself describing?
There appear to be reasons to accept both of the options. On one hand, if
fictions were abstract and eternal entities, whose existence is independent
of our conceiving of them, then how they are fictions would not be clear:
the same would go for numbers, geometric shapes, and laws of logic.
On the other hand, the possible worlds of which the modal realists and
ersatzists speak must have specific characteristics, such as maximality and
consistency. Even if they were artifacts, their authors would be constrained
in their creativity by the need for these characteristics.
Let us, at this point, clarify that the answer to the question about the
ontological status of the fiction depends partly on the specific version of F
we are considering. According to RF, for example, the fiction is not possible
worlds tout court, but possible worlds as described by Lewis (1986); given that,
for Lewis, there may be worlds that cannot be known by us, worlds would
have more in common with numbers or geometric shapes than with invented
characters such as Pinocchio. If, on the other hand, we endorse AF, then each
world is treated as a fiction of its own. At this point, Pinocchio turns out to be
a more useful point of comparison; but, we also lose—at least prima facie—
the option of claiming that there may be worlds that cannot be known by us.
That option is, nonetheless, regained, if we make reference to possibly existing
and yet unknowable fictions—a move that would appear to be countenanced

9781472524263_txt_print.indd 147 01/12/2015 09:44

 148 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

by Lewis’s modal realism. Modal realism would then be a fiction, according to

which there are infinitely many fictions, some of which we cannot know.
It would certainly be possible to argue that all works of literary fiction obey
certain rules of construction: for example, if we wanted to write a sonnet, we
would have to adhere to a certain metric schema. Of course, some genres,
such as the novel, are more liberal, but we could not write a one-page text
and legitimately call it a novel. The case of theories of possible worlds, then,
is not peculiar: if we accept that sonnets and novels are fictions, despite
the constraints on their creation, then we can also accept that a theory of
possible worlds is a fiction, despite its constraints.
Some theorists, such as Nolan (1997) and Kim (2005), have insisted that
the analogy between a possible-worlds theory and a novel seems to dissolve
if we consider the distinguishing trait of an artifact: the need for an author. A
novel has an author, but the fact that Foffo could have had milk and cookies
does not; the latter could even have happened without anyone having
conceived of it. The connection between fiction and reality is not analogous
to that between possibility and reality.
It would seem, however, that this variant of the fourth objection reflects
an incomplete understanding of modal fictionalism. The theory does not deny
that our modal sentences are true or false; therefore, for a modal fictionalist,
the fact that Foffo could have had milk and cookies does not depend on the
existence of someone who conceives of it. At the same time, the modal
fictionalist would claim that the explanation of that fact in accordance with
the fiction of a possible-worlds theory depends on the existence of an author.
A person does not have the power to do or undo modal facts, but she does
have the power to explain or not explain them. The fiction of which the modal
fictionalist speaks is the explanation itself, without which we cannot learn
what is or is not possible.
The metaphysical aspects of a fiction raise a fifth and final problem,
regarding the incompleteness of the descriptions we can give of a world.
In fact, the modal realist or the ersatzist can be exempt from the obligation
to provide a consistent and complete description of possible worlds, even
if they believe they exist; they can be so exempt insofar as they are able to
argue that worlds exist independently of those who speak of them. The modal
fictionalist, on the other hand, does not have this possibility: possible worlds
are fictions, just as a novel is; if a fiction is not described to the fullest, then
it is incomplete.
This is a delicate point for modal fictionalism. It is a tall order to provide a
consistent and complete description of all worlds; even more so if the fiction
is that of modal realism, according to which there could be worlds we will
never be able to imagine (see especially Woodward (2012) on this point).
Also relevant here is the interpretation of F under discussion—whether RF

9781472524263_txt_print.indd 148 01/12/2015 09:44

 MODAL FICTIONALISM AND MODAL AGNOSTICISM 149

or AF. In effect, it seems RF must admit that the modal realist’s fiction is
incomplete because it does not fully describe all the worlds of the theory.
A key point of modal realism is that it countenances more worlds than we
can represent (see Chapter 4), thus the fiction proposed by Lewis is such
that it cannot possibly describe all the worlds it suggests exist. Therefore,
the truth conditions of the sentences in our language can only be given in
an approximate manner, with the proviso that a consistent and complete
story of some worlds can be provided only (and at best) on a case-by-case
basis.
AF, by contrast, seems to have a more compelling solution. In fact,
according to this interpretation of F, worlds are fictions within the fiction of a
general theory of worlds: the latter fiction—the principal fiction—is not incom-
plete; what are incomplete are the fictions that it unfolds. Therefore, AF can
maintain that there is no incompleteness in the fiction that defines the modal
fictionalist theory, and that the incompleteness of the fictions of individual
worlds does not nullify the evaluation of the truth of a sentence because, in
accordance with what the modal realist or the ersatzist claim, such incom-
pleteness is irrelevant for the purposes of the evaluation.

§6.3. Modal agnosticism

One last theory of possibility that uses possible-worlds semantics remains to be
considered: modal agnosticism. As in the case of modal fictionalism, agnosticism
was developed in different philosophical realms: for example, in the philosophy
of science (see Van Fraassen (1980)) and in the philosophy of mathematics (for
a similar, but not strictly agnostic, approach, see Field (1980, 1989)). Articulate
defenses of an agnostic theory of modality are recent, and derive primarily from
an essay by John Divers (2004); see also Humberstone (2007).
Like modal fictionalism, modal agnosticism itself operates in relation to a
theory of possible worlds. As indicated by the term itself, modal agnosticism
gives no definitive answer regarding the truth of the theory it uses. Thus,
it shares with modal fictionalism a certain anti-realist strategy that aims to
exploit the theoretical strength of possible-worlds semantics—and possibly
of a full-blown theory of worlds such as modal realism—without committing
to the existence of worlds, and so without having to tell a sui generis story
about how we come to know modal truths. Indeed, Divers (2004) begins
with two observations regarding modal realism: that its analysis of modal
concepts is deeper than that of its rival theories; and that it commits us to
the existence of an infinite number of strange entities, that is, worlds just like
ours which, by definition, we cannot experience. The solution proposed, as in

9781472524263_txt_print.indd 149 01/12/2015 09:44

 150 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

the case of modal fictionalism, sets out to take advantage of the conceptual
analysis while, at the same time, avoiding a commitment to the existence of
worlds.
But this time, the idea is to deny assent to—to remain agnostic toward—
the sentences whose translation into modal realism (or another chosen
possible-worlds theory) commits us to the existence of worlds that are not
our own. This can be done in two ways, which give rise to two types of agnos-
ticism: strong and moderate.
Strong agnosticism denies assent to all sentences whose translation
into modal realism (or another possible-worlds theory) commits us to the
existence of worlds that are not our own. The moderate agnostic, on the other
hand, denies assent to only the sentences that commit us to the existence of
worlds whose existence we have reason to doubt (because we do not have
empirical proof of them, for example).
An analogy with the case of non-observable entities in physics can be
useful for clarifying this point. Let us suppose that a theory in physics
commits us to the existence of entities that are not directly observable, but
that behave in a way that is entirely similar to the behavior of molecules of
water; the strong agnostic would deny her assent to the theory in question,
while the moderate would offer it. The moderate agnostic, however, would
deny her assent to a theory that postulates the existence of non-observable
entities whose behavior is dissimilar to that of entities we are able to observe:
for example, he would deny assent to a theory according to which causal laws
between non-observable entities are not subject to any fundamental laws of
physics (e.g. conservation laws), but follow other sui generis laws instead.
The most attractive modal agnostic position—at least in terms of a theory
of modality—seems to be the moderate position. In fact, the strong modal
agnostic would deny assent to interpretations in modal realism (or another
possible-worlds theory) of almost any modal sentence; the only exclusion
would be the sentences that concern our world alone. The moderate modal
agnostic, on the other hand, would accept all sentences whose translation
into modal realism (or another possible-worlds theory) concerns situations
that are similar to those of our world. For example:

3 Foffo the cat could have eaten lava

is true if and only if:

4 There is a world in which Foffo the cat eats lava

is true. Since the cat, Foffo, and the lava mentioned in (4) are very much like
the cats and the lava found on our world, and since it does not seem contra-
dictory to believe (although it is unheard of) that Foffo, the cat, eats lava, the

9781472524263_txt_print.indd 150 01/12/2015 09:44

 MODAL FICTIONALISM AND MODAL AGNOSTICISM 151

moderate agnostic—unlike the strong agnostic—could accept (4) as true. But

if the sentence under evaluation were:

5 Donkeys can fly,

whose translation into modal realism is:

6 There is a world in which donkeys fly,

the moderate agnostic would suspend assent with regard to (6).

Although modal agnosticism superficially resembles the modal skepticism
introduced in Chapter 2, the difference between them should be clear: the
former accepts the plausibility of a theory of possibility—in particular, of
modal realism—but it denies assent regarding the truth of the modal asser-
tions in ordinary language whose translation into modal realism requires the
existence of situations that occur in worlds that are not our own and that are
entirely dissimilar to situations that occur in our world. It should be noted that
this position follows in the footsteps of the position defended in Mondadori
and Morton (1976), which tries to found the meaning of modal sentences on
what occurs and exists in the world in which we live; in the case of sentences
that do not have such a foundation, we are better off remaining uncom-
mitted. (We must point out, however, that Divers (2004) makes no mention
of Mondadori and Morton (1976).)
In light of what has been said, we can now introduce the concept of an
admissible sentence: for a modal agnostic, a sentence s is admissible if
and only if it is possible to express assent or dissent toward s. Hence the
following agnostic solution to the PP:

AGN: The sentence in natural language: “It is possible that: s” is true if

and only if: (i) its equivalent “s*” in modal realism (or another chosen
possible-worlds theory) is true; and (ii) “s*” is admissible.

It should be emphasized once again that AGN does not commit us to

accepting modal realism, but rather leaves open the question regarding which
possible-worlds theory is to be used. Ultimately, the agnostic simply accepts
what is proposed by a possible-worlds theory, within the limits of admis-
sible cases. The specific replies to the SPP, MPP, and EPP endorsed by the
modal agnostic will thus depend on which theory of possible worlds the modal
agnostic endorses, with a special proviso of admissibility for the EPP: we
cannot know of a modal fact unless the sentences describing it are regarded
as admissible by the specific version of modal agnosticism (moderate or
strong) that we have endorsed. But what counts as admissible? Is admissibility
an epistemic or metaphysical notion? We shall see in the next section that,

9781472524263_txt_print.indd 151 01/12/2015 09:44

 152 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

as innocent as it may appear at first, the proviso of admissibility, which limits

the range of sentences the modal agnostic is willing to regard as either true
or false, has important philosophical consequences for the modal agnostic.

§6.4. Modal agnosticism and possible worlds

We will examine two criticisms of modal agnosticism: the first is found in
Divers (2004); the second will be proposed here for the first time.
The plausibility of the modal agnostic proposal depends largely upon
that of point (ii) in AGN: only if there is a sufficiently large number of admis-
sible modal sentences will modal agnosticism be able to differentiate itself
from modal skepticism, reconcile the conceptual benefits of modal realism
(or another possible-worlds theory) with the refusal of its metaphysical costs,
and avoid having to rely on controversial entities such as fictions. But can the
agnostic supply a sufficient number of admissible sentences?
Let us take the following case as an example. Let us imagine a situation
in which a doctor has just developed a new pill, which we shall call p; the pill
is anticipated to defeat the HIV virus, but it has not yet been tested. It would
seem plausible to assert that, nonetheless, the following sentence is either
true or false:

7 The pill p can defeat the HIV virus.

The reason for which (7) is either true or false even before the test is
performed is that the test is not needed to instill certain capacities in the pill,
but rather to reveal the capacities the pill already has.
Nevertheless—it could be noted—the agnostic does not appear to have
the resources to assent or dissent to (7), and thus to say that it is either
true or false. That is, she could still express assent or denial to (7), but those
expressions would offer a thin bridge with truth and falsity. Thus, since there
are no situations in our world that are comparable to the one described in (7),
the sentence would not be admissible. Consequently, it would be neither
true nor false that the pill has certain capacities. (For a more complex but
analogous example, see Divers (2004: 681).)
The moderate modal agnostic has a reply to this criticism. In fact, the
problem raised does not appear to concern the metaphysics of possibility, but
rather its epistemology. And—the modal agnostic could insist—admissibility is
a metaphysical concept. The moderate modal agnostic, that is, could point out
that it is not true that there are no situations in our world that can be compared
to the one described in (7); there are, but, at the time of the test they have not
yet been realized and, therefore, the doctor in the example provided is not yet

9781472524263_txt_print.indd 152 01/12/2015 09:44

 MODAL FICTIONALISM AND MODAL AGNOSTICISM 153

aware of them. Hence, the problem is epistemological, not metaphysical: if

the admissibility of a sentence depends on that which occurs and exists in the
actual world—in its spatial and temporal totality—then in the actual world there
is a test and, therefore, there is a way to judge the truth of (7). However, in the
example given, the test has not yet been performed and, therefore, the doctor
would not know if (7) is true or false. But, this is simply an epistemic problem.
Let us suppose we were to change our example slightly by adding the
condition that the pill p was created with materials that do not exist in our
world and that cannot be obtained from materials of our world through any
process of manipulation whatsoever. At this point, (7) would arguably become
non-admissible, insofar as the test we are speaking of cannot be carried out
in our world. It should be noted, however, that the modified example has a
much lower argumentative appeal than the previous one: in the end, we have
good reason to doubt that the proposition expressed by a sentence that we
cannot test is either true or false.
For the strong modal agnostic, of course, the range of scenarios that
would be non-admissible is larger than for the moderate modal agnostic. It
includes all situations that do not obtain in our world, including, for instance,
the case of a pill that is created in our world but never tested: we cannot
truthfully attribute any causal power to such a pill, according to the strong
modal agnostic. Here, the modal intuition seems to be that the pill has causal
powers, even if they are never tested; thus, the strong modal agnostic pays
a higher price than does the moderate.
Let us now consider the second critical observation, which looks into
the specific relationship between modal agnosticism and modal realism.
The modal agnostic claims that it is possible to accept modal realism, even
without believing that there are worlds other than our own: the evaluation of
the truth of modal sentences is carried out by taking into consideration the
only world that we have reason to believe there is: ours.
Since we are moving within the context of modal realism, the modal
agnostic strategy must make use of counterpart theory; in particular, for
evaluating admissible sentences, the modal agnostic will find the relevant
counterpart in our very world (see Divers (2004)). For example, Foffo, the
cat, could have had milk and cookies because, in the actual world, there is
an individual who is similar to Foffo—let’s say, Fufi, the cat—who has milk
and cookies. In other words, according to modal agnosticism, a sentence
is admissible only if, given its translation in counterpart theory, in our world
there is at least one counterpart for each individual the sentence speaks of.
At this point, we might be asking ourselves whatever happened to the
worlds. It would appear that the modal agnostic proposes the following
interpretation of (possible) “worlds” in the language of her theory: all of
the worlds that, in the theory, are obtained through the simple recombination

9781472524263_txt_print.indd 153 01/12/2015 09:44

 154 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

of elements in the actual world, denote our world; all other worlds do not
denote it.
There are two problems with this interpretation. The first is that it exacer-
bates the objection that when we speak of what Foffo could have done, we
do not mean to speak of what Fufi is doing. The agnostic simply replaces
the individuals (of our world) of which modal sentences seem to speak with
other individuals (of the same world): yet, if we had wanted to speak of
certain individuals rather than of others, we would have. We did not mean to
speak of Fufi, but of Foffo.
The second problem is that modal agnosticism does not benefit from
the two principal conceptual advantages of modal realism. First, since all
admissible sentences concern worlds that are obtained through the simple
recombination of elements of the actual world, the modal agnostic could
never admit the existence of alien possibilities, as the modal realist does.
Secondly, since AGN makes reference to the modal concept of admissibility,
AGN does not reduce modalities, as modal realism aspires to do; more
simply, AGN explains them.

§6.5. Conclusions
The objective shared by modal fictionalism and modal agnosticism is to
borrow the philosophical strengths of modal realism (or of any other possible-
worlds theory that proves more philosophically palatable than others) while
avoiding its metaphysical costs. If achieved, the objective would deliver an
anti-realist paradise for modal theorists. Yet what emerges from our analysis
is that, for different reasons, both projects encounter numerous difficulties in
realizing their objective.
Modal realism, ersatzism, modal fictionalism, and modal agnosticism
exhaust the range of theories relying upon the existence of (possible)
worlds in order to solve the MPP and the EPP. We shall now turn to one last
theoretical option, which makes use of a new set of modal primitives in order
to deal with the truth or falsity of the propositions expressed by our modal
sentences: the so-called new actualism.

9781472524263_txt_print.indd 154 01/12/2015 09:44

 MODAL FICTIONALISM AND MODAL AGNOSTICISM 155

Study Questions
• What are the modal fictionalist’s solutions to the SPP, MPP, and EPP?
• What does it mean to be a fiction, according to the modal fictionalist?
• What types of things count as fictions, according to the modal fictionalist?
• What are the main objections that can be raised against modal
fictionalism?
• What is the distinguishing feature of modal agnosticism?
• What is the difference between strong and moderate modal agnosticism?
• What are the main objections that can be raised against modal
agnosticism?
• What are the common denominators of modal fictionalism and modal
agnosticism?

FURTHER READING
Modal fictionalism was first introduced by Rosen (1990). Some important
objections to the theory, and some possible modifications of it, were presented
shortly thereafter by Menzies and Pettit (1994) and Rosen himself (1995). For an
introductory discussion of modal fictionalism, see Nolan (2007); to situate the
fictionalist perspective within the broader context of fictionalism in philosophy,
see Kalderon (2005) and also Eklund (2007). The debate about fictionalism has
been lively and bears on several important details of the theory; among the many
discussion articles, see: Brock (1993); Rosen (1993, 1995); Noonan (1995); Hale
(1995a, 1995b); Divers (1995, 1999); Nolan and O’Leary Hawthorne (1996); Nolan
(1997); Baldwin (1998); Yablo (2001); Kim (2002, 2005); Dever (2003); Brogaard
(2006); Divers and Hagen (2006); Liggins (2008); Woodward (2008, 2011, 2012);
Sauchelli (2013); and Armour-Garb (2015).
Modal agnosticism is introduced by Divers (2004). A discussion of the specific
modal logic and modal semantics that best pair with modal agnosticism is found
in Humberstone (2007). An important paper discussing some ideas that are later
developed in modal agnosticism is Mondadori and Morton (1976).

9781472524263_txt_print.indd 155 01/12/2015 09:44

9781472524263_txt_print.indd 156 01/12/2015 09:44
 7
The new modal actualism

T he last place we will explore on our map of theories of modal metaphysics is

the so-called “new modal actualism”. Its proponents deny that our approach to
modality should start from necessity and possibility: they suggest that we start
instead from alternative notions—such as dispositions, essences, or other modal
properties—which we already have reason to include in our inventory of reality.
After an overview of the new modal actualism (§7.1.), two broad versions of the
view are discussed. The first is a version resting upon essences, which in turn
divides into individual essentialism and property essentialism (§7.2.). The second
version rests upon dispositions, and gives rise, in particular, to the so-called
dispositional theory of possibility (§7.3.).

*****

§7.1. The new modal actualism

T he last type of theory of possibility we will analyze covers some relatively

recent proposals that are not usually regarded as part of one and the
same theoretical perspective; following Vetter (2011a), we shall label this type
of theory new modal actualism. Vetter introduces new actualism as follows:

These new actualists, as I shall call them, do not feel the onus of providing
an actualist account of possible worlds. Possible worlds, they say, may be
a useful formal device in modal logic (as well as in other formal contexts),

9781472524263_txt_print.indd 157 01/12/2015 09:44

 158 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

but they have little to do with the metaphysics of modality. Instead of

accounting for possible worlds, then, these theorists seek to provide an
account of modality directly; their shared aim is to identify, within the
actual world, the grounds, source or truthmaker of modal truths.
—Vetter (2011a: 742)

All the proposals falling under the label of new modal actualism share two
important tenets. First, these theories deny that possible worlds provide a
viable analysis of modal sentences. For this reason, Contessa (2010) suggests
a different label for these proposals: hardcore modal actualism. The idea is
that the soft-core modal actualists contend that the actual world is ontologi-
cally privileged with respect to all other worlds, while making use of possible
worlds to make sense of modal sentences; the hard-core modal actualists, on
the other hand, aim to provide an analysis of modal sentences that does away
with possible worlds. The new modal actualism thus shares with the views
discussed in Chapter 2 and Chapter 3 the rejection of possible worlds. It is
the second tenet that distinguishes the new modal actualism from the views
discussed in those early chapters. New modal actualists aim to substitute talk
of possible worlds with talk of modalities that are possessed by individuals,
such as essences, dispositions, or other modal properties. In so doing, and
unlike the theories we discussed in the forementioned chapters, the new
modal actualists offer a cognitivist analysis of modality.
By now, our reader will have noticed that different theories of possibility
approach their task by privileging one or two of the three sub-problems of
the problem of possibility (PP). Those, let us recall, are the semantic problem
of possibility (SPP), the metaphysical problem of possibility (MPP), and the
epistemic problem of possibility (EPP):

PP: What does it take for a certain situation to be possible?

SPP: What does it mean to say that a certain situation is possible?
EPP: How do we come to know that which is possible?
MPP: What sort of entity is a possible entity (of any given kind—a
possible individual, property, state of affairs, or …)?

As Vetter remarks in the passage cited earlier, the new modal actualism
seems to pay special attention to the MPP. In particular, the new modal
actualism calls into question the way that the MPP has been discussed
for the past fifty years. It has been standard in quantified modal logic (QML)
to privilege the two modalities of necessity and possibility—represented, as
we have seen, by the box and the diamond symbols. Theories of possibility
have been driven to solve the MPP by postulating the existence of possible
worlds just insofar as they have conformed with this semantic decision.

9781472524263_txt_print.indd 158 01/12/2015 09:44

 THE NEW MODAL ACTUALISM 159

Possible worlds suit the two modalities of necessity and possibility and are
in accordance with the semantics suggested to complement both QML
and counterpart theory; although, as we have seen (Chapters 4 to 6), it is
very difficult to provide a sound theory of possibility in terms of possible
worlds. Thus, for the past fifty years, discussion of the MPP has been largely
dominated by the view that, if modal sentences do express content, they are
about possible worlds—whatever these turn out to be.
The new modal actualist denies that our approach to the PP should
start from necessity and possibility: instead, she suggests that we start
from alternative notions—such as dispositions, essences, or other modal
properties—which we already have reason to include into our inventory of
reality. Thus, instead of talking about possible worlds, the new modal actualist
will carry forward a theoretical analysis of possibility founded on one (or
more) of these alternative notions; necessity and possibility will have to be
explained in terms of those notions. While the new actualism stakes its claim
to a place on the map of theories of possibility on a metaphysical point, it is
nonetheless important to stress that arguments for versions of that approach
can depend on broader reasons, having to do with achieving a suitable philo-
sophical explanation of modal sentences or with semantic considerations.
We shall see an example of an appeal to semantic considerations in the new
essentialist’s insistence on the foundational explanatory role of essences,
What is the alleged theoretical advantage of doing away with necessity
and possibility as the core concepts by means of which to analyze modal
sentences? Different new modal actualist views will provide different
answers to this question. In general, we can point to two advantages. First,
modal notions such as dispositions or essences are specific to individuals,
thus they allow for some finer explanatory distinctions. A notable example
was provided by Kit Fine: by invoking a modal theory that uses the notion of
essence, we can easily explain why Socrates is intuitively more fundamental
than the singleton (a set with exactly one element) that contains Socrates as
its sole element; according to Fine, this is because it is part of the essence
of Socrates’s singleton to have as its sole element Socrates, while it is not
part of the essence of Socrates that there is a singleton of Socrates. No
such account is open to the theories based on possible worlds discussed in
Chapters 4 to 6, because allegedly Socrates and its singleton exist in exactly
the same worlds. Of course, supporters of theories based on possible worlds
can try to devise some explanation, based for instance on impossible worlds;
nonetheless, their view will pay a theoretical price to accommodate cases
that Fine’s theory does not have to pay, because it makes use of essences.
The second theoretical advantage of new modal actualism is to simplify
the philosophical explanation of modal sentences by having it draw from the
philosophical explanation of other philosophical issues that involve modalities.

9781472524263_txt_print.indd 159 01/12/2015 09:44

 160 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

For instance, dispositionalism starts from the observation that we already

need dispositions in order to account for the ascription of certain properties
and powers to actual entities; if this is the case, then why not try to employ
dispositions to explain modal sentences involving other types of scenarios
as well? Similarly, Fine’s view starts by pointing out the need for essences
in explaining important metaphysical differences, such as that between
Socrates and its singleton; once this is acknowledged, we may then try to
employ essences to shed light on other types of scenarios as well.
Our discussion of new modal actualism will concentrate on two types of
theories: the new actualist essentialism (§7.2.) and dispositionalism (§7.3.).
The new actualist essentialism encompasses at least two different views:
one—championed by Fine and labeled individual essentialism—that is more
attentive to the essences of necessary existents, such as Socrates and its
singleton; and another—championed by Michael Jubien and Brian Ellis and
labeled property essentialism—that views essences as types of properties.
The two views share some common ground, and may perhaps be developed
into a unified view in the years to come; also for this reason, we shall treat them
together. Dispositionalism, for its part, moves from the widespread underlying
intuition that the world is imbued with powers, capacities, and dispositions;
the view was partially anticipated by Mondadori and Morton (1976) and Martin
and Heil (1999); it has been more recently developed by Pruss (2002), Borghini
(2003), Borghini and Williams (2008), and Contessa (2010).
Before moving on, however, we should recall another theory that some
readers may associate with new modal actualism: the proposal put forward
by Timothy Williamson in a series of recent writings—including those of
2005, 2007b, and 2010—according to which the concept of what is possible
is subject to the concept of the counterfactual situation; in other words, the
response to the PP proposed by Williamson plays upon the ability of human
beings to imagine counterfactual situations: possibility is simply a concept
derived from the expansion and systematization of this ability. Because
of Williamson’s insistence on the idea that all individuals are necessary
existents, we shall discuss this view in the next and final chapter—Chapter 8.

§7.2. Individual essentialism and

property essentialism
We have encountered essentialism twice already: the first time was in
Chapter 4, while discussing a metaphysical variant of modalism inspired
by the work of Jonathan Lowe; the second time was in Chapter 6, when
considering Alvin Plantinga’s ersatzist view, according to which worlds are

9781472524263_txt_print.indd 160 01/12/2015 09:44

 THE NEW MODAL ACTUALISM 161

states of affairs constituted by individual essences and their properties

(including essential properties). These positions differ from one another both
in their understanding of essences and in the roles that essences play in the
explanation of modal sentences. According to Lowe, an essence is the real
definition of an individual; Foffo, the cat, is identical to its essence. For this
reason, Lowe’s proposal can provide an adequate foundation for modalism.
According to the modalist, truths in QML must be accepted without recurring
to possible worlds, and positing the existence of essences (as real defini-
tions) helps us found modal truths of QML on what is actually the case. On
the other hand, Plantinga’s essences are part and parcel of possible worlds,
conceived as states of affairs. Plantinga’s position seems compatible with
the view that essences are real definitions of individuals and are identical to
them; however, for Plantinga there are some essences that are not actual,
and yet exist as constituents of possible worlds. In his view, then, essences
found the truth of sentences of possible-worlds semantics, which the
modalist refuses to accept as true.
We shall discuss two versions of the new actualist essentialism: individual
essentialism and property essentialism. These share two commitments. For
one, essences—and not necessity and possibility—are the starting point
in explaining matters concerning modalities; second, essences provide a
suitable metaphysical terrain to propose a novel semantics and epistemology
of modality. That said, the two views differ significantly with respect to the
theory of essence they support. Thus, while it is reasonable to think that a
unified new actualist essentialist view may be proposed, such a view would
require adjusting some important details of its ingredient theories. Let us
discuss each view, in order.
Individual essentialism shares a great deal with Lowe’s understanding of
essences; indeed, Lowe explicitly cites Fine as a source of inspiration for
his work. As we suggested in Chapter 4, Lowe’s project can be regarded as
part and parcel of modalism, only to the extent that essences are invoked
to provide the metaphysical foundation of the language of QML. However,
if essences act as the foundation for a more updated semantic theory of
modal sentences that proposes itself as an alternative and sui generis view
of modality, then what we have is no longer a modalist view. Because
Fine’s proposal aims to explain modality in terms of the essences of actual
individuals, it fits the new actualist essentialist tag.
Fine defended individual essentialism in a number of papers, including
those published in 1994 and 1995. His argument moved from a polemic
against the view that essence is a modal notion:

It is my aim […] to show that the contemporary assimilation of essence

to modality is fundamentally misguided and that, as a consequence, the

9781472524263_txt_print.indd 161 01/12/2015 09:44

 162 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

corresponding conception of metaphysics should be given up. It is not

my view that the modal account fails to capture anything which might
reasonably be called a concept of essence. My point, rather, is that the
notion of essence, which is of central importance to the metaphysics of
identity is not to be understood in modal terms or even to be regarded as
extensionally equivalent to a modal notion.
—Fine (1994: 3)

For Fine, explaining metaphysical necessity requires, not possible worlds, but
rather identity. (For an argument showing that the notion of essence cannot be
captured in terms of possible—or the combination of possible and impossible—
worlds, see Torza (2015).) The proper understanding of identity delivers a theory
of essences, according to which an essence defines an entity; such a definition
can then act as the explanatory foundation for metaphysically necessary truths.
We have described the workings of a similar doctrine in Chapter 4, and the
reader can refer back to that discussion for more detail on the underlying
conception of essence. Fine’s doctrine licenses the following answer to the PP:

IE: The sentence in natural language: “It is possible that: s” expresses

the existence of an individual whose essence grounds the truth of that
sentence.

It is important to clarify an aspect of Fine’s view that will be discussed again

in Chapter 8: according to Fine, there are different varieties of necessity
depending on the domain of discourse.

[E]ach class of objects, be they concepts or individuals or entities of some

other kind, will give rise to its own domain of necessary truths, the truths
which flow from the nature of the objects in question. The metaphysically
necessary truths can then be identified with the propositions which are
true in virtue of the nature of all objects1 whatever.
—Fine (1994: 9)

Essences found metaphysical truths; logical truths, instead, are founded

upon logical necessity, which does not rely on a doctrine of essences; scien-
tific truths rely on nomic necessity, which again has sui generis foundation.
Thus, individual essentialism and the theory of essences that accompanies it
do not provide a full-fledged account of all modal talk; more modestly, they
provide an answer to the MPP for those modal claims that do indeed concern

1
What Fine calls “objects” here is what we refer to in this book as “individuals”, and for the sake
of consistency, we stick to the terminology previously adopted in the volume.

9781472524263_txt_print.indd 162 01/12/2015 09:44

 THE NEW MODAL ACTUALISM 163

metaphysical truths. It is in this sense that individual essentialism differs from

the version of modalism that is inspired by Lowe’s doctrine of essences: the
latter aimed to provide a general theory of modality grounded on essences,
while individual essentialism suggests a piecemeal approach.
The second version of the new actualist essentialism is property essen-
tialism. According to this theory, the solution to the MPP is rooted in the
essences of the properties of actual entities. The details of the theory have
been articulated in two ways, which we will label, respectively, intrinsic
essentialism and scientific essentialism. Most notably defended by Jubien
(2007, 2009), intrinsic essentialism argues that the essences of properties
depend upon the intrinsic character of properties; thus, if Foffo is a cat, Foffo
is essentially an animal because it is the intrinsic character of the property
Being a cat to entail the property Being an animal; this notion of intrinsic
character, moreover, founds the possibility that Foffo is black. Thus, Foffo’s
property of Being a cat, in virtue of its essence, founds many (possibly
infinite) metaphysical truths about Foffo.
Scientific essentialism, on the other hand, sees the roots of metaphysical
necessity in the fundamental properties discovered by the natural sciences.
It is necessary that salt dissolves in water because the property Being a
molecule of salt and the property Being a molecule of water, in virtue of their
essences, bear a certain necessary relationship to each other; such necessary
relations found the laws of nature concerning water and salt. Analogously, all
metaphysical truths are founded upon the relationship that properties bear
to each other in virtue of the essential character of fundamental properties.
Brian Ellis (2001) and Alexander Bird (2007), who develop ideas found in
Shoemaker (1980), are among the most notable supporters of this view,
which has received considerable attention in contemporary metaphysics of
science.
Finally, we can condense the two articulations of property essentialism as
follows:

PE: The sentence in natural language: “It is possible that: s” expresses

the existence of a property whose essence grounds the truth of that
sentence.

Individual essentialism and property essentialism rest on two quite different

readings of the doctrine of essences. Indeed, Fine has openly criticized
scientific essentialism (e.g. 2002) and he would probably not endorse intrinsic
essentialism either. Still, the two versions of the new actualist essentialism
do share an important aspect: they both reverse the order of metaphysical
explanation, contending that, at the metaphysical level, necessity and possi-
bility are explained in terms of what holds true of essences (whatever these

9781472524263_txt_print.indd 163 01/12/2015 09:44

 164 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

turn out to be). The two versions are relatively new to the literature on the
metaphysics of possibility and we may one day see a theory that conjoins
individual and property essentialism, properly revised—a core new actualist
essentialist view.
Our discussion of the new actual essentialism illustrates a characteristic
trait of the new modal actualism tout court. Over the past fifty years, theories
of possibility have attempted to provide a unified philosophical analysis of
all modal sentences that is cohesive and systematic; such analyses have
ultimately rested on a cohesive semantic picture, be it specified in terms
of the more familiar possible-worlds semantics, counterpart theory, or a
simplified modalist account. The new modal actualism proposes a different
methodology, rooted in a piecemeal explanation of modal sentences. As
attractive as it may seem at the semantic level, the cohesive picture also
required bizarre solutions to the MPP and the EPP. The new modal actualism
proposes local but more palatable solutions to the three sub-problems of
possibility, with particular attention to the MPP.

§7.3. Dispositionalism
Let us now turn to dispositionalism. We are all accustomed to attributing
certain abilities to people, animals, things, and events: Giovanni can ride 8m
waves; Foffo, the cat, can climb up to the top of the cherry tree in a total of
5 seconds; the magnificent Etruscan statue that we saw at the museum of
Tarquinia is very fragile (and thus able to break); the sea storm is so strong
that it could swallow up the largest ship of the fleet. And we have a name
for these abilities: dispositions (or, as we will also call them, dispositional
properties); in attributing abilities to people, animals, things, or events, we
are attributing dispositions to them. The tendency in ordinary language is to
consider dispositions to be properties of the entities they are attributed to:
just as we say that Giovanni lives in Volterra and measures 1.8m in height, so,
too, we say that he can ride 8m waves. To live in Volterra, to have a height
of 1.8m, and to be able to ride 8m waves are all properties of Giovanni.
Dispositionalism adopts this tendency as its own, and sets out to include
among the properties of the actual world certain dispositional properties;
the dispositionalist theory, then, explains the meaning of modal sentences
of ordinary language in terms of the dispositions of the entities of the actual
world.
The idea is simple to put into words, but a bit more complex to develop.
The exposition of the theory will be divided into two parts here: the first will
show that, in order to explain the attribution of certain abilities to Giovanni,

9781472524263_txt_print.indd 164 01/12/2015 09:44

 THE NEW MODAL ACTUALISM 165

we must postulate the existence of dispositional properties; the second will

show that dispositional properties are enough to explain the meaning of
modal sentences.

§7.3.1. Dispositions
What are dispositions? First and foremost, they are properties that are
attributed to the entities of the world. They are properties with a particular
characteristic: they do not manifest themselves constantly, but only
occasionally—and, in some cases never. It is entirely natural to say that the
Etruscan statue is very fragile: and yet, the statue has never manifested its
fragility—and we hope it never will. We are the ones who have attributed
that property to the statue, based on our observation of other individuals
composed of the same material. For this reason, dispositional properties are
distinguished from categorical properties: the latter properties are always
active, so to speak, provided they are possessed by an individual. For a more
in-depth discussion, see Mumford (1998: 64–92).
This particular trait of dispositions is responsible for their aura of philo-
sophical diffidence in empiricist circles, which is deeply rooted in the work
of Hume and Ryle (1949). In simple terms, empiricists do not believe in
the existence of that which does not manifest itself through a sensory
experience; and dispositions are, by definition, entities that can be hidden
from our senses. An empiricist would not admit that the Etruscan statue
has the property of being fragile: of course, we say that it is fragile based
on our observation of other individuals composed of similar material; but this
is only a projection of our minds, just like when we begin a role game by
declaring that the salt shaker is the wife and the sugar bowl is the jealous
husband.
A more detailed description of the empiricist strategy is as follows: every
sentence that contains one or more dispositional terms (terms that seem to
refer to an ability or dispositional property) must be interpreted as a sentence
in a conditional form (such as: “If …, then …”) in which the dispositional
terms are eliminated. The truth of the conditional sentence is then evaluated
based on the relevant evidence that we have accumulated in the past.
Let us take an example. Let us once again consider the Etruscan statue in
the museum in Tarquinia; we can say that:

1 The Etruscan statue at the entrance of the Etruscan museum in

Tarquinia is very fragile.

An empiricist would interpret (1) in the following manner:

9781472524263_txt_print.indd 165 01/12/2015 09:44

 166 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

2 If the Etruscan statue at the entrance of the Etruscan museum in

Tarquinia were subjected to even a feeble blow, then it would be
shattered.

Whether what is stated in (2) is true will be established based on certain

observations made regarding the behavior of individuals comparable to the
Etruscan statue, both in structure and in material. Accordingly, there is no
need to postulate the existence of the dispositional properties of the statue;
we are the ones who project upon it certain conditions.
Nonetheless, there is an entire philosophical tradition that has made use
of dispositions, beginning with Aristotle. And it is in light of this tradition
that, starting with the works of Harré (1970), Mellor (1974), and Mondadori
and Morton (1976), dispositions have been rehabilitated within the realm of
analytic philosophy. Today, there are numerous theories that postulate the
existence of dispositions, namely Martin (1994), Ellis (2001, 2002), Molnar
(2003), and Heil (2003). Philosophers such as Karl Popper have made use of
them as well (1990).
The chief motive behind this blossoming interest is the vast spectrum of
philosophical questions that dispositions can help to answer; in fact, they
have been used in the philosophy of mind, for example, to explain certain
mental attitudes (such as belief or desire); in the philosophy of science, they
have been used to explain what a law of nature is; in economics and rational
choice theory, they have been used to explain the behavior of agents.
Yet, insofar as it denies the existence of entities that can only be partially
experienced, such as dispositions, the empiricist theory has the merit of
simplifying metaphysics. What, then, are the reasons to admit the existence
of dispositions?
There are ultimately two reasons, and they have been brought to light in
the recent debate mainly by—among others—Martin (1994), Mumford (1998),
Molnar (2003), Schrenk (2009), Mumford and Anjum (2011), and Contessa
(2013). The first has to do with the interpretive strategy: attempting to explain
a dispositional term based on certain conditions is often a hollow exercise.
Sentence (2), for instance, is not a correct interpretation of (1): it is not
enough to mention a blow, feeble as it may be, in order to specify the condi-
tions under which the statue would break. The blow’s impact is a necessary
condition, but it is not sufficient. If the statue were to be encased in solid
cellophane, it would not shatter, even if it were flung to the ground; if certain
laws that govern the interaction among the particles that compose the statue
were to change while it were flung to the ground, it would not shatter.
The empiricist would respond that it would suffice to add a clause in which it
is assumed that the environmental conditions (physical, chemical, atmospheric,
and so on) remain the same (these are the so-called ceteris paribus clauses):

9781472524263_txt_print.indd 166 01/12/2015 09:44

 THE NEW MODAL ACTUALISM 167

3 Provided that the environmental conditions remain the same, if the

Etruscan statue at the entrance of the Etruscan museum in Tarquinia
were subjected to an even feeble blow, then it would be shattered.

The theorist of dispositions would not be satisfied with this solution, however.
Asserting that the actual conditions must remain the same does not explain
which necessary conditions cause the statue to be fragile; consequently, it
does not explain what fragility consists in. As noted by Carnap (1936), the
sufficient conditions that allow an individual to manifest a disposition can be
manifold. We might consider, for example, the conditions that cause a vineyard
to be able to produce a good wine: certain temperature values, atmospheric
conditions, the presence or absence of organisms in the environment … Who
could list them all? It would seem more reasonable to assert that the vineyard
has the ability—that is, a certain disposition—to produce good wine.
Furthermore, there is at least one additional problem for the conditional
analysis: what type of relationship subsists between the antecedent and
the consequent of the conditional statement expressed in (2) and (3)? We
are certainly not dealing with a simple conditional statement such as “If …
then”, as (2) and (3) suggest; in fact, if this were so, it would be enough for
the antecedent to not be satisfied to make the conditional statement true.
A more accurate analysis must therefore be produced; in order to do so, we
must use a modal notion—for example, we could affirm that: in the case in
which the conditions remain the same, if the antecedent were to be true,
then the consequent would follow necessarily. In order to explain the modal
notion that has been introduced, the empiricist must turn to one of the
theories previously illustrated; and, if she aspires to find a satisfying interpre-
tation, she will no longer have to rely solely upon past observations, but also
on certain merely possible situations.
Let us now turn to the second difficulty encountered by an empiricist
approach to dispositions—and second reason to admit their existence. It
is based on a simple observation: it would seem reasonable to assert that
some dispositions, though they exist, do not manifest themselves. Let us
once again consider the statue. Fortunately, it has never been broken. But
let us suppose that the museum curator has secretly been tempted on
multiple occasions to melt it in the old furnace nearby (with all due respect
to the curator!) One night, falling prey to an unusually strong temptation, the
curator starts the furnace and tosses the statue among the flames, and it
slowly melts without breaking. The statue has never been broken and never
will break. Nevertheless, it seems plausible to remain of the opinion that the
statue was fragile: how can an empiricist justify this opinion?
There are, fundamentally, two solutions, both of which are problematic.
The first, attempted by Quine (1974), is based, in simple terms, on the fact

9781472524263_txt_print.indd 167 01/12/2015 09:44

 168 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

that the statue is an individual of the same type, in structure and materials, as
individuals that have broken in the past; it is, therefore, legitimate to attribute
fragility to the statue. However, the relevance of resemblance cannot be
taken as a primitive fact: why do we assume that a resemblance in structure
and materials will be accompanied by a resemblance in behavior? In order to
provide an adequate answer, we must refer to a modal notion—that of natural
law, for example. And so we find ourselves once again dealing with a modality.
The second, perhaps more persuasive, but certainly more bizarre, solution
was proposed by David Lewis. For the modal realist, it is sufficient for a world
to exist in which a statue that is particularly similar to the one at the entrance
of the Etruscan museum in Tarquinia breaks in order to verify that the statue
in our world could have broken.
This solution is not satisfying, however. As demonstrated first by Martin
(1994) and Bird (1998), it cannot be applied to cases in which the intrinsic
properties of individuals change. Let us suppose that, after years of research,
a factory successfully produces a material that, when it is not subjected to
the kind of blow that would be fatal for a statue like the one in the museum,
is fragile; if, however, the material is subjected to such an impact, it goes from
being fragile to indestructible; as soon as the force ceases, it becomes fragile
once again. Suppose we make a new statue from this material. How could
we appeal to possible worlds to justify our belief that the statue is fragile?
In the end, given that the disposition to go from fragile to indestructible is
intrinsic to the statue, no possible world exists in which it does not possess
this property. It should be noted that denying that it is possible to make
a statue like the one described amounts to forgoing one of the virtues of
modal realism: its practically unlimited ability to represent possible scenarios.
The modal realist, therefore, is able to explain the dispositions that do not
manifest themselves, but not those that change.
In sum, the empiricist analysis of sentences that contain dispositional
terms proves insufficient, both because it is unable to provide an adequate
translation of sentences in terms of conditional sentences, and because it
is unable to account for the possibility that the dispositions of an individual
change through time. For these reasons, many have deemed it necessary to
include dispositions among the entities of the world.

§7.3.2. The dispositional theory of possibility

The idea at the heart of dispositionalism ensues from what has just been
concluded: if we have reasons that are not ad hoc to postulate that there
are properties with a modal character, then why not attempt to explain
metaphysical modality in general using these properties?

9781472524263_txt_print.indd 168 01/12/2015 09:44

 THE NEW MODAL ACTUALISM 169

The first thing to note about the dispositionalist proposal is its methodo-
logical approach. Again, the order of metaphysical explanation is reversed with
respect to the theories we examined in Chapters 4 to 6: the metaphysical
possibility and necessity of certain situations is explained in terms of the
dispositions possessed by actual entities. While the possible-worlds theories
are prompted by a theory of the meaning of modal sentences (e.g. possible-
worlds semantics), dispositionalism is inspired by certain metaphysical
considerations: the theoretical opportunity to include dispositions among the
entities of our world. Dispositionalism therefore takes its cue from its solution
to the MPP, that is, its explanation of what a possible entity is. According to
the dispositionalist, the entities whose possibility or necessity we express are
situations; what we are left with is the following principle (see also Borghini
and Williams (2008) for a more thorough discussion of P):

P: A situation s is possible if and only if there is at least a disposition

whose manifestation includes s.

It should be noted that P does not require dispositions to be manifested:

all that is needed is their existence. Of course, if they were to manifest
themselves, all the better. But, as we explained previously, one of the
reasons to include dispositions among the entities of the world is precisely
the fact that they could exist without manifesting themselves; therefore, we
cannot eliminate them in favor of their manifestations. The statue can break
precisely because it has the disposition to break under certain circumstances;
it does not need to manifest its disposition in order for that disposition to be
attributed to it; and we must postulate the existence of the disposition since
we do not have a better alternative for explaining that the statue can break.
Another distinctive feature of the theory is that a single disposition
can be used to account for various possibilities. For example, the fragility
of the statue would also account for the possibility of a scandal occurring at
the museum, or of the collection of contemporary Etruscan works losing one
of its most cherished pieces.
Finally, the dispositionalist draws upon the metaphysical theory P in order
to respond to the more general PP in the following manner:

D: The sentence in natural language: “It is possible that: s” expresses

the existence of one or more dispositions whose manifestation
includes the situation expressed by “s”.

It should also be pointed out that, for the dispositionalist, the EPP remains in
the background, as is the case for many of the theories we have considered.
Typically, a division of roles is invoked. How to come to know of the existence

9781472524263_txt_print.indd 169 01/12/2015 09:44

 170 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

of the dispositions that make a given modal sentence true is a task that
applies not only to the dispositionalist philosopher, but—as a matter of
principle—to all other members of society as well.

§7.3.3. Two problems for dispositionalism

Although dispositionalism was formulated only recently, as we have noted,
it is rooted in a long tradition of philosophical work that makes use of dispo-
sitions. For this reason alone, the view is vulnerable to certain established
objections. We will focus on two here. Other critical remarks and objections
have been raised recently by Vance (2014), Wang (2015), Yates (2015), and
Austin (2015).
The first objection concerns the number of possibilities dispositionalism is
able to account for. Among the motivations that drive the dispositionalist to
begin with the MPP when considering modality is eliminating the ontological
commitment to the existence of an infinity of possible worlds. The dispo-
sitionalist maintains that there is only one world, our world, in which there
is an infinity of dispositions. However, this could come off as a theoretical
limitation. Are dispositions enough to account for the immense variety of
possibilities there appear to be in our world?
This is certainly a delicate point of the theory. Fortunately, the disposi-
tionalist is able to account for a greater number of possibilities than there
might initially appear to be. In fact, it should be noted that, if manifested,
each disposition brings into being the existence of other dispositions. In
other words: with the passage of time, the entities of the world manifest
some of their dispositions, from which more and more dispositions arise.
For example, Giovanni can ride 8m waves with great skill; if manifested, this
disposition gives rise to Giovanni’s additional disposition of being among the
athletes who have been summoned for the next world surfing championship;
if manifested, this disposition gives rise to Giovanni’s additional disposition of
advancing to the next round of the world surfing championship; and so on and
so forth.
In light of this observation, we are able to distinguish various types of
dispositions. We give the name first-degree dispositions to those dispositions
that an entity (or a collectivity of entities) is able to manifest simply if the right
environmental conditions present themselves (that is: without the entity, in
turn, having to change). We give the name second-degree dispositions to
those dispositions that an entity is able to manifest if and only if a first-degree
disposition manifests itself; we give the name n +1 degree dispositions to
those dispositions which an entity is able to manifest if and only if an n degree
disposition manifests itself.

9781472524263_txt_print.indd 170 01/12/2015 09:44

 THE NEW MODAL ACTUALISM 171

The dispositionalist would argue that her theory includes all dispositions
of all degrees. Consequently, there would also be dispositions that are far
from manifesting themselves, as they describe scenarios which could have
been realized if a whole other series of occurrences had been realized, though
it never was. For example, Foffo could have eaten oysters in Oristano on
December 13, 2013, since there is an n degree disposition to do so that could
have manifested itself if a whole series of other dispositions had manifested
itself. This is how the dispositionalist is able to account for remote possibilities
such as: that dinosaurs had never been extinct, that life had never developed
in the universe, or that Rino Gaetano had given a concert in Bulgaria in 2003.
Yet, this result may not be enough to answer the criticism. The modal
realist, in particular, would point her finger at alien possibilities: certainly the
dispositionalist would be able to account for all of the possibilities that have
to do with the way in which our world could have been. But why not also
imagine that there are genuine possibilities that are not ways in which our
world could have been?
At best, the dispositionalist could reply that our world—itself—could have
developed in different ways from the start, giving rise to worlds that are a far
cry from the one in which we live. Yet, she must admit that there is no way
to include the existence of possibilities that our world has not been, is not,
and will never be, to any degree, inclined to manifest.
The second objection to dispositionalism concerns the very nature of
dispositions. (For more on this point, see Borghini (2009).) A disposition is
always defined in terms of a certain situation whose manifestation it can
contribute to. It would almost appear that a situation contains the seeds
of something that does not exist, something that it intends to bring to
completion, if certain conditions were to present themselves. Some have
identified in this scenario two related problems: first, it would appear that
dispositions have an intentional character, similar to our thoughts; second, it
would appear that this character compels us to admit the existence of the
situations that dispositions intend to bring to completion. Nevertheless, it
would seem absurd to attribute intentions to the Etruscan statue; and, if the
dispositionalist were to admit that situations have intentions, then she would
also have to admit the existence of possible worlds.
There is at least one way to respond to this criticism. In fact, the two
problems that it raises originate from a single, erroneous temptation: to define
dispositions. The dispositionalist would insist that the concept of disposition
is primitive: if we could define it in terms of certain situations, we would.
In reality, speaking of intentional situations and characters gives us nothing
more than a partial and metaphorical explanation of what a disposition is. It is
metaphorical because the concept of intention would not be applied to dispo-
sitions in the same way we apply it to a mind: it is not associated with natural

9781472524263_txt_print.indd 171 01/12/2015 09:44

 172 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

language, and much less with the idea of free will. It is partial because many
dispositions cannot be fully understood through the description of certain
situations.
At the beginning of the book, we alluded to the theory that that which
is, is what it is in virtue of that which it could be. In other words: actual and
possible existence are strictly and indivisibly connected. To some extent,
dispositionalism is founded on this theory. It is able to naturally account for
the way we seem to conceive of that which surrounds us. A chair is such as
it is because it has the disposition to carry out a certain function for human
beings; a gene is such as it is because it has the disposition to carry out a
certain function in the development of an organism; a piece of material is gold
because it has the disposition to melt at a certain temperature, or to become
worn with time much more slowly compared to other materials; Giovanni was
elected as chief executive not only in virtue of what he did, but also in virtue
of what he can do. Certainly, things can also be seen differently. But the
interpretation offered by dispositionalism is compelling. We need only admit
that, among the entities of our world, there are also certain properties that,
as a matter of principle, are not always observable. Considering the complex
and obscure theoretical commitment of possible worlds, this does not seem
to be an absurd price to pay.

Study Questions
• What are the two distinctive aspects of the new modal actualism?
• What is individual essentialism?
• What is intrinsic essentialism?
• What is scientific essentialism?
• Why do individual essentialism and property essentialism seem
incompatible?
• What is a disposition?
• What characterizes an empiricist approach to dispositions?
• What is the relationship between dispositions and conditionals?
• How is possibility explained, according to the dispositionalist?
• What is the degree of a disposition and why is it relevant for the
dispositionalist?

9781472524263_txt_print.indd 172 01/12/2015 09:44

 THE NEW MODAL ACTUALISM 173

FURTHER READING
The new actualism is a recent development of the debate on modality and it is not
systematically treated in introductory texts on the topic. Useful first readings are
Vetter (2011a and, especially 2015).
A good starting point to appreciate the theory of essence underlying individual
essentialism is Lowe (2012). The classic texts that outline the position are by
Fine (1994, 1995); for a broader perspective, see also the papers collected in
Fine (2005). Fine’s view has been widely discussed in recent years and expanded
into the so-called “theory of metaphysical grounding”, which will be discussed in
Chapter 8; for an overview of this latter theory, see Bliss and Trogdon (2014).
Intrinsic essentialism is defended in Jubien (2007, 2009). Scientific
essentialism has a connection with the views on metaphysical necessity proposed
by Putnam and Kripke in the 1970s; for a classic exposition of the theory, see Ellis
and Lierse (1994) and Ellis (2001, 2002); an important contribution was given by
the more recent Bird (2005, 2007); for some background texts, see especially
Shoemaker (1980) and Swoyer (1982). For some early criticisms of the theory, see
Bealer (1987); for a more updated discussion, see Kistler (2002), Anderson (2005),
and Corry (2011).
For an introduction to the theory of dispositions, see the classic by Mumford
(1998). Among the works that defended a form of realism regarding dispositions,
see Harré (1970), Mellor (1974), Popper (1990), Martin (1994), Ellis (2001, 2002),
and Molnar (2003). For the relationship between dispositions and intentionality,
see Borghini (2009). A wider and ongoing debate regards the role of conditionals
in defining dispositions; for some recent contributions, see: Kment (2006); Manley
and Wasserman (2008, 2011); Schrenk (2009, 2010); Vetter (2011b, 2013a, 2013b,
2014); and Nolan (2015).
Among the articles that predate dispositionalism and anticipate some of its
ideas, see Mondadori and Morton (1976), Martin and Heil (1999), and Pruss (2002).
The view was first proposed in Borghini (2003), and then in Borghini and Williams
(2008), Jacobs (2010), Mumford and Anjum (2011), and Vetter (2015). A wide
debate has more recently ensued, which contains important developments; see
especially, Contessa (2010), Vance (2014), Austin (2015), Wang (2015), and Yates
(2015).

9781472524263_txt_print.indd 173 01/12/2015 09:44

9781472524263_txt_print.indd 174 01/12/2015 09:44
 8
Necessities, necessary
existents, and their bounds

W hile possibility and necessity are logically and, therefore, theoretically

entangled with one another, certain questions pertaining to the truth of
necessary statements deserve to be discussed independently of specific theories
of possibility. First of all, a more thorough analysis of the different varieties of
necessity (e.g. logical, natural, metaphysical, deontic, normative) is in order (§8.1.).
Then, by means of the theory of grounding, an array of truths is analyzed, typically
concerning mathematical entities, such as geometrical figures and numbers, as
well as abstract entities such as sets and moral values (§8.2.). Next on the agenda
is the theory of necessary existents proposed by Timothy Williamson (§8.3.).
Completing the picture is a discussion of two views that challenge the boundaries
of—respectively—what seems to exist and what seems to be possible: the first
view is Meinongianism (§8.4.), according to which there are things that do not
exist; the second view is dialetheism, according to which some sentences are
such that, at the same time, they are true, and their negations are true, as well
(§8.5.). At the very end, an epilogue reviews the main principles discussed in the
volume, and draws two general conclusions regarding the philosophical analysis
of modal talk (§8.6.).

*****

S o far we have been especially concerned with theories of possibility and,

in particular, with their metaphysical underpinnings. Necessity has been
kept in the background. It is now time to confront the relevant questions
concerning necessity, in order to deliver a well-rounded picture of the
metaphysics of alethic modalities.

9781472524263_txt_print.indd 175 01/12/2015 09:44

 176 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

As we explained in the Introduction, alethic modalities concern both that

which is possible, and that which is necessary. Possibility and necessity
are most often regarded as twin concepts: to say of a certain situation that
it is necessary amounts to saying that it is not possible that it does not
obtain; conversely, to say of a certain situation that it is possible amounts
to saying that it is not necessary that it does not obtain. Nonetheless, when
elaborating a theory of modality, it is standard to discuss possible scenarios,
leaving to the side the issue of how necessity claims shall be adjudicated.
An underlying presumption may be that, once matters of possibility have
been settled, matters of necessity will follow suit. For instance, a theorist
endorsing the existence of possible worlds may work under the presumption
that what is true at all possible worlds is what is necessary: thus, for that
theorist necessity is spelled out in terms of the boundaries of the possible.
Nonetheless, claims of necessity deserve separate treatment. This is
because there is controversy over where to draw the boundaries of possi-
bility and, once we start delving into this issue, sui generis difficulties arise.
The aim of this chapter is to introduce us to these difficulties and point out
the primary ways of addressing them. To begin, there seem to be different
varieties of necessity. We have been focusing especially on metaphysical
necessity, but this modal notion is also often employed to convey the
lack of an alternative from other points of view, such as: logic (logical
necessity); natural science (nomic necessity); duty (deontic necessity);
laws and norms (normative necessity); or simply pragmatic considerations.
A proper discussion of the varieties of necessity will therefore be in order
(§8.1.). Second, this book has seen several mentions of theories of essences.
In recent years, much attention has been devoted to the special relationships
of dependence between entities that essentially depend on each other; the
most important outcome of this strain of research is the so-called theory of
grounding. Essential dependence most strikingly concerns mathematical
entities, such as geometrical figures and numbers, and abstract entities, such
as sets; but, to the extent that essences are attributed to concrete entities—
such as chairs, trees, and people—a theory of essential dependence will
help clarify claims about them as well. So we shall discuss the theory of
grounding (§8.2.). Next on the agenda is the theory of necessary existents—
of those entities that exist out of necessity, regardless of what scenario we
are entertaining. Timothy Williamson has famously defended the view that
every individual exists out of necessity; we shall discuss Williamson’s view
and explain how it sits with respect to the theories of possibility covered in
the previous chapters (§8.3.). We shall complete our treatment of necessity
by discussing two views that challenge the boundaries of—respectively—
what seems to exist and what seems to be possible: the first view is
Meinongianism (§8.4.), according to which there are things that do not exist;

9781472524263_txt_print.indd 176 01/12/2015 09:44

 NECESSITIES, NECESSARY EXISTENTS, AND THEIR BOUNDS 177

the second view is dialetheism, according to which some sentences are

such that, at the same time, they are true and their negations are true as well
(§8.5.).

§8.1. The varieties of necessity

In the previous chapters of this book, we have studied the problem of possi-
bility and its cognate semantic, epistemic, and metaphysical problems. It is
now time to consider a parallel set of questions pertaining the other classic
alethic modality: necessity.

PN: What does it take for a certain situation to be necessary?

SPN: What does it mean to say that a certain situation is necessary?
EPN: How do we come to know that which is necessary?
MPN: What sort of entity is a necessary entity (of any given kind—a
possible individual, property, state of affairs, or …)?

Clearly, the answers to the problems of possibility we have considered

translate also to the problems of necessity. Typically, for instance, a linguistic
ersatzist will tend to provide a linguistic ersatzist account of necessity,
regarding a necessity claim as true when there is a complete and consistent
description of a world according to which it is true (cf. Chapter 5); or, a modal
fictionalist will regard a necessity claim as true when, once we pretend that
modal realism is true, the claim comes out as true (cf. Chapter 6). Yet the
problems of necessity bring to light more vividly an issue that has loomed
large in our discussion of modality, but that we have not directly addressed:
the classification of alethic modalities into two sole notions, possibility and
necessity, falls short of fulfilling the theoretical and explanatory roles that
we need. Theories of modality are employed to address issues that concern
different sorts of entities, from numbers to ordinary material objects to the
entities postulated by the natural sciences; it would not be surprising if,
depending on the sort of entity under consideration, a different modal notion
would apply. Hence, it would not be surprising if it were to turn out that
our notions of mathematical necessity, metaphysical necessity, and natural
necessity were distinct from one another.
Kit Fine has prominently defended the importance of recognizing a variety
of independent modal notions, primarily necessities:

The earlier literature on modality, arising from the work of Quine, was charac-
terized by an unwarranted contempt for modal notions. The subsequent

9781472524263_txt_print.indd 177 01/12/2015 09:44

 178 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

literature, arising from the work of Kripke, has been characterized by an

unwarranted enthusiasm. This enthusiasm has taken two different, though
related, forms. The first, which we may call “modal mania”, is a matter of
seeing everything as modal; every notion which is somehow associated
with modal features is itself taken to be modal. The second, which we
may call “modal myopia”, is a matter of seeing all modality as metaphysical;
every modality is somehow to be understood as a form of metaphysical
modality.
—Fine (2005: 9)

We have encountered an instance of “modal mania” in our discussion of

essentialism (see Chapter 7). Fine’s and Lowe’s proposal to regard essence
as a non-modal notion stood in contrast to the modal–maniac perspective
(prompted by the work of Kripke) that defines essences in terms of possible
worlds. We shall return to “modal mania” in the next section, where we will
consider alternative takes on the notion of grounding, some of which are
modal and some non-modal. For the time being, though, we are going to
concentrate on the challenge faced by the “modal myopic”: Is there really a
plausible way to subsume all sorts of necessity under one sort of necessity
and, likewise, all sorts of possibility under one sort of possibility? Let us
call modal monism the view that gives a positive answer to this question.
The modal monist, that is, will contend that—say—metaphysical necessity
and possibility suffice as basic notions, in terms of which to account for
all modalities. On the other hand, let us call modal pluralism the view that
gives a negative answer to the question. Modal Pluralism comes in various
forms: from those who hold that there are only two or three fundamental
modalities to more liberal positions. We shall consider some examples. For
now, we should note that under the modal pluralist umbrella falls also the
view recently defended by Mumford and Anjum (2011), according to which
dispositional modality is a third modal notion, not fully analyzable in terms of
possibility or necessity (see also Chapter 7, where this view was recalled in
connection to the dispositional theory of possibility).
It is first important to point out the intuition—and ensuing theoretical
desideratum—motivating modal monism. Start from our most unrestricted
sense of possibility: according to it, all sorts of things are possible for Foffo,
the cat. Foffo could grow wings and start flying; Foffo could turn into a mango
juice, but for five minutes alone; and so on. Now, take all the things that Foffo
could be: this fixes the range of what is metaphysically possible for Foffo; by
doing so, we will also be making explicit what is metaphysically necessary
of Foffo (this may include its being a cat, or its having a certain origin, or,
more modestly, its bearing a certain name). The modal monist intuition is
that any other notion of alethic possibility that applies to Foffo is obtainable

9781472524263_txt_print.indd 178 01/12/2015 09:44

 NECESSITIES, NECESSARY EXISTENTS, AND THEIR BOUNDS 179

by appropriately restricting the range of scenarios capturing what Foffo could

be; conversely, any other notion of alethic necessity that applies to Foffo is
obtainable by appropriately enlarging the range of features that Foffo must
possess.
For instance, it is an alleged biological necessity that Foffo evolved from
wild cats, because (as far as we know) members of the species Felis catus
were domesticated members of the species Felis silvestris; such necessity can
be explained in terms of the driving intuition by simply adding the requirement
of being domesticated from Felis silvestris to the list of necessary features
that metaphysically characterize Foffo. On the other hand, while it is biologi-
cally necessary that Foffo evolved from wild cats, one could play with the idea
that it is allegedly not a physical necessity that Foffo evolved from wild cats:
the physical molecules that compose Foffo now could have been arranged as
they are independently of any other previous arrangement of particles (e.g.
a lab scientist could have arranged them in that way). So, among the physi-
cally necessary features of Foffo we do not find evolutionary features; but we
do find physical features, such as being made of electrons: arguably, Foffo
must be made of electrons, according to physics. This is how we can tell
that physical necessity is a restriction of metaphysical necessity: Being made
of electrons is a physically necessary feature of Foffo, but not a metaphysi-
cally necessary feature (Foffo could be made of some particle alien to our
world, for instance); at the same time, there are no metaphysically necessary
features that are not also physically necessary features of Foffo: thus, every
scenario that is physically possible for Foffo is also metaphysically possible for
Foffo. An equal line of reasoning applies to the relationship between biological
modalities and metaphysical modalities. Hence, the driving intuition of modal
monism leads to the theoretical desideratum that alethic modalities can all
be understood in terms of one sort of modality, typically metaphysical modal-
ities. Ideally, alethic modalities would even be hierarchically nested: at the top
we find metaphysical modalities, followed by—in order—physical, chemical,
biological, … All modalities, in the end, would be understood as varieties of
the metaphysical sort. (On the relationship between natural and metaphysical
necessity, see also Sidelle (2002, 1989).)
Modal monism descends from the Kripkean idea that necessity, essence,
and identity are captured in terms of truth at all possible worlds. By restricting
or enlarging the number of worlds that we consider possible, we can envisage
different notions of necessity, essence, and identity. Ideally, there will be one
unrestricted domain of worlds, encompassing all possibilities and fixing the
most fundamental truths about necessity, essence, and identity. If other
domains of discourse employing modalities can be hierarchically arranged,
from the most general to the most specific, we can then envisage these as
nested restrictions of that initial domain.

9781472524263_txt_print.indd 179 01/12/2015 09:44

 180 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

For a Quinian modal skeptic, modal monism does not even get started:
the Quinian would stop the monist right at the outset, arguing that we cannot
intelligibly speak of the necessary features of an entity, such as Foffo (cf.
Chapter 2). For this reason, the Quinian and the modal monist—as noted by
Fine in the foregoing passage—stand at opposite ends of a spectrum. The
position defended by Fine is intermediate: it is a form of modal pluralism.
Modal pluralism can be advocated more on the basis of general methodo-
logical considerations than by moving from specific arguments. It is indeed
not easy to see how a scenario could be naturally possible—that is to say,
physically, biologically, or chemically possible—and yet metaphysically impos-
sible, though perhaps some taxonomic puzzles, such as the platypus, or
some strains of bacteria may suggest nice examples (cfr. respectively Eco
(1999) and Ereschefski (2010)). But, the general methodological point is that
metaphysics is an autonomous domain of inquiry with respect to—say—
physics, chemistry, or biology. It makes little sense to try to understand,
much less explain, metaphysical necessities and possibilities in terms of
natural necessities and possibilities. Arguably, there is more to metaphysics
than what the natural sciences contemplate, while there may be modal
truths in the natural sciences that are hard to square with some metaphysical
tenets, as taxonomic puzzles indicate. We are better off thinking of the
domains as autonomous, though of course we can ask from time to time,
when it seems opportune, whether the domains are related or what sorts of
theoretical incompatibilities they might generate.
Fine’s view is a form of sparse pluralism, because according to him there
are only three distinct and autonomous sources of necessity: metaphysical,
natural, and normative. All other sorts of necessity and possibility can be
accounted for in terms of one of those three. We can envisage much less
parsimonious forms of pluralism—forms of abundant pluralism—which distin-
guish between dozens of varieties of modality.
We shall not venture into the details of abundant pluralism. Let us
nonetheless consider the modest plurality of forms of modality introduced
by the distinctive nature of mathematical and logical truths. Such truths
seem indeed to enjoy a special status with respect to spatio-temporal events
and facts. Consider first a material entity—say Foffo, the cat—and take the
following sentence expressing an alleged necessary truth regarding Foffo:

1 Necessarily: Foffo is a cat.

One way to render the content expressed by sentence (1) is to see

it as suggesting that, whatever is the case, Foffo is a cat. The alleged
truth expressed by (1), that is, explains how Foffo is connected to all
worldly phenomena: Foffo can play the cat role, and only that role, in

9781472524263_txt_print.indd 180 01/12/2015 09:44

 NECESSITIES, NECESSARY EXISTENTS, AND THEIR BOUNDS 181

those phenomena. By contrast, consider a sentence expressing an alleged

necessary mathematical truth:

2 Necessarily: 2 + 2 = 4.

This time, it seems that the content expressed by (2) is not about the
necessary role that numbers play in the workings of worldly phenomena.
More than informing us that “2 + 2 = 4” is true whatever the case, (2) is
telling us that “2 + 2 = 4” no matter what is the case. The necessity of
mathematical truths, that is, seems to rest on their independence from what
is the case, rather than on a fixed role in what is the case (as in Foffo being
a cat). Similar considerations apply to logical truths and, perhaps, truths in
certain other fields (such as music theory). The distinction between the two
ways of being necessarily true—having a fixed role in worldly phenomena
versus being independently true—is sometimes rendered by claiming that
mathematical and logical truths are transcendent. Regardless of the specific
terminology, it is evident that there are good reasons to consider the source
of necessity for mathematical and logical truth distinct from the source of
necessity for natural and, perhaps, metaphysical truths. Modal monists face
the challenge of finding one unitary source for such disparate modalities.
It is opportune to recall that this volume is confined to alethic modalities.
Modal monism, as described here, is the contention that all alethic modalities
can be subsumed under one set of modalities. As we discussed in the
Introduction, there are several additional modalities besides the alethic,
including the temporal, the deontic, and the epistemic. A far more radical
form of monism would therefore contend that any modality—not only the
alethic ones—can be subsumed under one set of modalities. Typically, such a
position is suitable only for those who believe that all there is descends from
a single principle or entity. For instance, if God is ultimately the source of all
that there is and all truths—including all necessary and possible existents and
all necessary and possible truths—then perhaps we have reason to endorse
a form of radical modal monism; or, consider the position of those who
believe that we cannot escape language—that all there is is but a linguistic
construction: in this case, too, there may be room to endorse a radical form
of modal monism.
Before bringing this section to a close, let us note one form of necessity
that is at times subsumed under alethic necessity: normative necessity.
Some people contend that what is good can ultimately be explained in terms
of laws of nature (that is, certain natural necessities) or in terms of certain
metaphysical necessities. The positions that defend either of these claims are
far too numerous and articulated to be discussed here. But it is important to
recall the link between alethic modalities and normative ones.

9781472524263_txt_print.indd 181 01/12/2015 09:44

 182 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Bringing into focus necessity, rather than possibility, has allowed us to

start recognizing a complex universe of modalities—a universe about which,
in principle, each of the theories discussed in the previous chapters has
something to say. We shall now turn to some specific notions and problems
of the debate regarding necessary entities and truths, before returning at the
very end of the chapter to the intersection between theories of possibility and
necessity.

§8.2. Grounding
We shall now examine the so-called notion of grounding, with a particular
focus on its relationship to necessity. We will start with some examples of
grounding. Consider the following sentences:

3 The singleton of the singleton of Foffo exists because the singleton of

Foffo exists.

4 When Luca calls out “Foffo!”, Foffo comes to Luca, because Luca
always called it “Foffo”.

5 Foffo is a good cat because Foffo likes to play.

These sentences illustrate three varieties of grounding. Sentence (3)

exemplifies ties typical of mathematical or logical entities: the hierarchy of
sets that has Foffo as an initial element (Foffo, the singleton of Foffo, the
singleton of the singleton of Foffo, …) is grounded in Foffo. Were Foffo
not to exist, the hierarchy would not exist. And yet, because Foffo exists
the hierarchy exists also, and so does any specific relationship between its
elements. The key feature—at least for our purposes—of the relationship
between the singleton of the singleton of Foffo and the singleton of Foffo is
that the relationship cannot be captured simply in terms of possible worlds.
For any world where the singleton of Foffo exists, the singleton of the
singleton of Foffo also exists; and vice versa. And yet, the singleton of Foffo is
arguably more fundamental than the singleton of the singleton of Foffo. Such
fundamentality is captured precisely by saying that the singleton of Foffo
grounds the singleton of the singleton of Foffo.
Sentence (4) exemplifies the semantic ties between words and worldly
phenomena or entities. There is an element of contingency in having called
Foffo, the cat, “Foffo”; and yet, Luca’s consistent use of the name throughout
Foffo’s life is grounded in the fact that “Foffo” names Foffo. Such a grounding
tie is not causal: by naming Foffo, the cat, “Foffo” Luca did not really cause
any of his subsequent speech acts (at best we may say he somehow

9781472524263_txt_print.indd 182 01/12/2015 09:44

 NECESSITIES, NECESSARY EXISTENTS, AND THEIR BOUNDS 183

influenced them). Moreover, semantic grounding ties cannot be captured

in terms of possible worlds: for any world in which Foffo, the cat, exists,
Foffo is “Foffo”. Thus, the semantic tie between Foffo and its name cannot
be captured in terms of either causation or possible worlds: we introduce
grounding as a special notion standing for such a tie. Notice that the element
of contingency regarding the fact that Foffo, the cat, was named “Foffo” by
Luca is likewise unable to be explained in terms of possible worlds: if Foffo
bears its name necessarily, I cannot really depict a world were Foffo is not
named “Foffo”, because that would be a world where Foffo is not Foffo.
Such semantic contingency is indeed best explained in terms of a semantic
notion of necessity, understood as distinct from the metaphysical notion of
necessity that fixes the identity of Foffo.
Sentence (5) exemplifies theoretical ties between different facts. The fact
that Foffo likes to play grounds the fact that Foffo is a good cat. The relationship
between the two facts is arguably not causal: Foffo’s playfulness may tend
to generate fondness in those who encounter Foffo, but it seems prudent
to keep separate such a psychological reaction from a judgment regarding
whether Foffo is a good cat or not. The relationship between the two facts
cannot be captured in terms of metaphysical necessity—not unless we
grant the controversial assumption that facts about a cat’s playfulness
metaphysically necessitate facts about its goodness. What might instead be
operative here is some other notion of necessity, probably of a normative
variety.
Sentences (3), (4), and (5) indicate the range of grounding ties. The
sampling by no means shows the full spectrum of grounding ties: for
instance, grounding can be employed to depict the ties between laws, insti-
tutional facts, moral facts, professional standards, just to name a few. Such
an impressive spectrum explains why it is disputed whether “grounding”
picks out a single notion or several distinct ones. We find here an element of
analogy with the discussion of the varieties of alethic necessity: on one hand
it would be desirable to subsume a number of case studies under a single
notion; on the other, once we try to do that, we incur the risk of leaving key
subtleties out of our analysis, so that we end up with an unsatisfactory result.
The examples set by (3), (4), and (5) demonstrate the need for an account
of grounding that is nuanced enough to capture the different ties conveyed
in each sentence. If a single notion of grounding is to pick out ties between
mathematical entities, semantic links, and theoretical claims about cats’
goodness, it must possess an amazing flexibility.
It may not have escaped notice that I referred to grounding ties, and not to
grounding relations. This is because it is important to remain neutral, at this
preliminary stage, regarding the sort of entity involved in grounding. Both early
logical analysis of grounding, such as Fine (2001), as well as notable, more

9781472524263_txt_print.indd 183 01/12/2015 09:44

 184 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

recent developments, such as Correia (2010) and Fine (2012), treat grounding
as a sentential connective, leaving somewhat open its metaphysical details.
This seems a sensible move, especially if we aim to have a notion that is
able to pick out a diverse range of scenarios. Thus, it may turn out that,
on some occasions grounding concerns facts, on others, propositions, on
others still, material objects or other individual entities; and yet, despite such
metaphysical differences, the same notion is at play.
Our brief foray into grounding is meant especially to emphasize the alleged
inadequacy of alethic necessity—in particular when understood as truth in all
possible worlds—for capturing absolute or universal ties. Semantic ties, the
nexuses between sets within a hierarchy, and the links between theoretical
facts cannot be readily accommodated in possible-worlds semantics. For this
reason, grounding may be regarded as a topic that reveals the shortcomings
of modal monism.
That said, further inspection may reveal that the cases assembled under
the label of grounding do not involve modal notions after all, or at least
not alethic modalities. Scenarios displaying grounding would likely then fall
into more than one non-modal category. For instance, some examples of
grounding may concern identity, which may be argued to be a non-modal
notion; other examples may involve normative necessity, which—as we saw
in the previous section—may well be autonomous with respect to alethic
necessity; other instances may involve semantic necessity, which again is
arguably independent of the alethic; and so on. So, modal monism may be
salvaged, if we are able to show that instances of grounding do not call upon
alethic modalities.
One important lesson for our discussion of alethic modalities can be
learned from considering the problem of grounding. At their most ambitious,
some modal monists (including prominent ones, such as Lewis) hoped to
explain even the problem of grounding itself in terms of alethic modalities;
those among them who were possible-worlds theorists and modal realists
hoped to reduce all modal talk to the non-modal existence of certain entities in
certain worlds. The past few years have cast severe doubts on the aspirations
of modal monists. More than fifty years have gone by since the formulation of
possible-worlds semantics; the semantics has proved very useful in a number
of cases; yet it seems far from providing sufficient conceptual resources to
analyze even the primary cases of grounding. If we add that the reductivist
efforts of modal realism also face formidable challenges, as we saw in
Chapter 4, we can conclude that it is important to take a piecemeal approach
to modal talk.

9781472524263_txt_print.indd 184 01/12/2015 09:44

 NECESSITIES, NECESSARY EXISTENTS, AND THEIR BOUNDS 185

§8.3. Necessary existents

Next on our agenda is a discussion of so-called necessary existents. This is a
longstanding issue that occupied philosophers such as Fine and Plantinga in
the 1980s (see Fine’s “Postscript” to Prior and Fine (1977), Plantinga (1983,
1985), and Fine (1985)). In more recent years, Linsky and Zalta (1994, 1996)
and most notably Williamson revived the debate. Here we shall focus
especially on Williamson’s work; in the Further Reading section, the reader
will find additional bibliographical information regarding classic texts on this
topic, direct replies to Williamson, and recent contributions to the debate.
In 2013, Williamson published an influential study on the logic and metaphysics
of modality titled Modal Logic as Metaphysics. The volume developed and
refined some of the central ideas Williamson had presented in previous works,
including those of 1990, 1998, 2000, 2002, and 2007b. A controversial idea
emerging from Williamson’s writings is that the domain of individuals inhab-
iting different possible worlds is fixed: any individual that inhabits a world is a
necessary existent, that is, it exists in all worlds. Foffo, the cat, for instance,
is—according to Williamson—a necessary existent: that is, Foffo cannot fail to
exist, despite any modal intuition to the contrary we may have.
Williamson’s argument for the theoretical need to posit necessary existents
is strikingly simple and flies in the face of our most basic modal intuitions, as
he himself notes:

It seems obvious that I could have failed to exist. My parents could easily
never have met, in which case I should never have been conceived and
born. The like applies to everyone. More generally, it seems plausible
that whatever exists in space and time could have failed to exist. Events
could have taken an utterly different course. Our existence, like most
other aspects of our lives, appears frighteningly contingent. It is therefore
surprising that there is a proof of my necessary existence, a proof that
generalizes to everything whatsoever.
—Williamson (2002: 233)

The alleged “proof” comes in five sentences.

(A) Necessarily, if I do not exist then the proposition that I do not exist
is true.

(B) Necessarily, if the proposition that I do not exist is true then the
proposition that I do not exist exists.

(C) Necessarily, if the proposition that I do not exist exists then I exist.

9781472524263_txt_print.indd 185 01/12/2015 09:44

 186 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

(D) Hence, necessarily, if I do not exist then I exist.

(E) Necessarily, I exist.

Let us briefly assess the argument, illustrating some ways in which its contro-
versial assumptions can be resisted (for further discussion, see the Further
Reading section). Sentence (A) establishes a necessary link between a
negative scenario (viz. that I do not exist) and the existence of a proposition.
Notoriously, philosophers have sought to deny that negative sentences, such
as “I do not exist”, require the existence of negative states of affairs or facts in
order to be regarded as true. If I do not exist, then I am featured in no states
of affairs or facts, and there is therefore no proposition featuring me. Thus, a
first way to resist the argument is to deny that my nonexistence necessitates
the existence of a proposition regarding me because there is no proposition
whatsoever where I am featured. That is, let us concede that my nonexistence
necessitates the existence of a proposition; that proposition, nonetheless, is
about entities other than me (one could maintain, for instance, that it is about
a total fact, where I am not featured). Whether all negative sentences can be
explained-away in terms of non-negative states of affairs or facts is, however,
controversial. Thus, sentence (A) has at the very least some plausibility.
Sentence (B) lays out a thesis about the metaphysics of propositions: true
propositions must exist. A way to resist such a thesis is to argue that not all
true propositions must exist: some of them can be metaphysically reduced to
other propositions. In the case in question, the proposition that I do not exist
can be metaphysically reduced—say—to a proposition about the totality of
existents, in which I do not feature. But such a move would require consid-
erable work on the metaphysics of propositions.
Sentence (C) exemplifies the thesis that every individual featured in a true
proposition must exist. Strategies to resist this claim have been endorsed in
contemporary metaphysics by conventionalists, for example, who wish to
maintain that propositions about things such as tables and chairs are true,
while denying that there are tables and chairs (see Hirsh (1997, 2009) and
Varzi (2002) for some examples).
Sentence (D) follows from sentences (A) through (C), and cannot be
questioned unless we give up basic inference rules of logic. The same goes
for sentence (E), even though it employs the more controversial logical
principle that, from a contradiction, anything follows; dialetheists, whom we
shall discuss in section §8.5., reject this principle and would therefore deny
that, because sentence (D) states a contradiction, sentence (E) follows.
Williamson’s argument, whether sound or not, puts in evidence a major
issue in modality, especially in connection with possible-worlds semantics:
our conception of propositions is deeply entrenched with the analytic

9781472524263_txt_print.indd 186 01/12/2015 09:44

 NECESSITIES, NECESSARY EXISTENTS, AND THEIR BOUNDS 187

results of our modal theory. This is a vivid example of the interplay between
logic, semantics, and metaphysics: we end up with a metaphysical claim
concerning necessary existents based on some theses concerning proposi-
tions and truth. Specifically, if we endorse the three main theses displayed
in sentences (A) through (C) of Williamson’s argument, then we end up with
a fixed domain of entities for all possible worlds. While Williamson goes on
to construct a full-fledged modal theory based on this result, other authors
would regard the result as a major problem because it overrules—rather than
accommodating—modal intuitions as basic as that I might not have existed.
(For a comparison of Williamson’s and Lewis’s modal theories that is particu-
larly relevant to the ends of this volume, see Divers (2014d).)

§8.4. Meinongianism
Are there entities that do not exist? This question, which is about the
boundaries of reality, probes our ideas regarding the limits of possibility, and
therefore also regarding what is necessarily the case. It is thus important
to address this question in the context of a discussion of necessity and its
boundaries. So-called Meinongians answer the question positively. They have
enjoyed a poor reputation in logic and metaphysics; but, building on the works
of authors such as Parsons (1979, 1980) and Routley (1980), today, they are
having a comeback (see Nelson (2012) and Priest (2005) for an overview). In
this section, we shall focus especially on how Meinongianism is entangled
with the debates on the metaphysics of possibility and necessity.
Meinongianism—which owes its name to Alexius Meinong, an Austrian
philosopher and psychologist renowned for his theory of non-existent entities
(see Marek (2013))—is motivated by some simple and commonplace intui-
tions. Consider Foffo, the cat. Thus far, we have presumed that Foffo actually
does exist; but, we must confess, Foffo does not actually exist: Foffo is a
creature of our imagination. Now, in a certain sense, even though Foffo,
the cat, is not of this world, it still makes sense to maintain that Foffo does
exist; after all, we have been talking about this cat for an entire volume,
illustrating complex theories of modality thanks to it. Assuming that at least
some content was passed along through our illustrations involving Foffo, such
illustrations must have been about something. Foffo is no different than the
entities we encounter in novels, stories, movies, or in our imagination, and
that do not actually exist: to the extent that we interact (even only at the level
of imagination) with them, we may want to concede that they exist. That is
what the Meinongian does: according to the Meinongian, Foffo, along with all
other intentional entities, does exist.

9781472524263_txt_print.indd 187 01/12/2015 09:44

 188 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

Admitting the existence of non-actually existent entities stretches the

bounds of possibility. David Lewis’s modal realism seems to give in to a form
of Meinongianism and was indeed accused of being Meinongian shortly after
being presented (see Linsky (1991)). If to be a Meinongian means to accept
the existence of non-actual entities, then Lewis certainly fits the description.
For Lewis, there are infinite worlds and “actuality” is an indexical expression:
there is no world such that it is true, in that world, that all existing entities are
actual. Meinongianism—understood as the thesis that there are entities that
are non-actual—therefore turns out to be necessarily true in modal realism.
The affinity between Lewisian modal realism and Meinongianism also helps
to explain why modal realism is theoretically more powerful than rival theories
of possibility on the market: in a sense, Lewis’s theory makes room for the
existence of an infinity of non-existents, thereby expanding the bounds of
possibility.
Nonetheless, contemporary Meinongians would regard Lewis’s modal
realism as a spooky version of Meinongianism. This is because Lewis’s
worlds are complete and consistent: for any property and any individual inhab-
iting any world, it is determinate whether that individual has that property at
that world. But, our starting intuition was much simpler than that. It seemed
to require a much more naïve principle, which is taken to characterize
so-called naïve Meinongianism:

Naïve Principle: For any condition on entities, there is a unique entity

satisfying exactly that condition.

The condition, here, stands for any assortment of properties; the requirement
of uniqueness is to ensure that the same specific entity is picked out on any
occasion—for example, that when we speak of Foffo we speak always of the
same entity.
Naïve Meinongianism faces several problems, some of which nicely illus-
trate the link between Meinongianism and theories of alethic modality. Two of
these problems deal precisely with completeness and consistency, the two
requirements that we find in modal realism (as well as in all other theories of
possibility we have surveyed so far) but that are not mentioned in the Naïve
Principle. First, consider incompleteness. According to the Naïve Principle,
there is exactly one entity that exemplifies the condition Being Foffo the cat.
That entity has only one property—Being Foffo the cat; anything else about
that entity is indeterminate: it is neither true nor false that it is white, neither
true nor false that it ate fish and potatoes, and so on. Most philosophers dislike
incomplete entities because they violate the so-called principle of bivalence,
according to which any sentence (expressing a proposition) is either true or
false. According to the view under discussion, there are infinite sentences

9781472524263_txt_print.indd 188 01/12/2015 09:44

 NECESSITIES, NECESSARY EXISTENTS, AND THEIR BOUNDS 189

about the entity that exemplifies the condition Being Foffo the cat that are
neither true nor false. Is it really possible that entities violating bivalence exist?
The Naïve Principle violates not only bivalence, but also the principle
of non-contradiction (contradictory sentences cannot both be true at the
same time). According to the Naïve Principle, there is exactly one entity
that exemplifies the condition Being Foffo the cat and Being better than
any cat. But, if so, then this entity is better than itself (since it’s better
than any cat). Not only that, but according to the Naïve Principle there is also
exactly one entity that exemplifies the condition Being Foffo the cat and Being
self-distinct. This entity, then, is and is not identical to Foffo, the cat. Is it really
possible that entities violating consistency exist?
The third problem (already raised by Russell in response to Meinong;
see Nelson (2012)) is that, according to Naïve Meinongianism, there is no
restriction on the conditions on entities. We may require as a condition that
the entity exist. So, according to the Naïve Principle, there is exactly one
entity that exemplifies the condition Being Foffo the cat and existing. We
started from the intuition that there is a certain Foffo the cat that we have
been talking about and that does not exist, and Naïve Meinongianism has
forced us to admit its existence. This is striking and seems just plain wrong.
Meinongians have come up with different strategies to cope with the
difficulties of the Naïve Principle. One such strategy, famously defended by
Parsons (1980), distinguishes between nuclear and extra-nuclear properties:
the nuclear properties alone fix the identity (or, as some people say, the
“nature”) of an entity, while the extra-nuclear do not. The alleged violations
of bivalence and non-contradiction concern extra-nuclear properties and thus
are not really violations of bivalence and non-contradiction, after all; as for
existence, it is regarded as an extra-nuclear property, so it cannot be added
to the condition regarding Foffo, the cat. A second strategy (see Zalta (1988)
and (1983)) distinguishes between exemplifying and encoding a property:
exemplification is the familiar relation between entities and properties,
while encoding is the special relation at work in the case of conditions on
intentional entities, such as Foffo, the cat. According to this strategy, the
Naïve Principle specifies only encoded properties; but, we shall not worry if
encoded properties violate bivalence or contradiction, or if they violate intui-
tions about existence: since it is not exemplification to be at stake, no real
theoretical problem is posed. A third strategy (see Priest (2005) and Berto
(2013a) and (2012)) goes in the direction of Lewisian Meinongianism and is
known as Modal Meinongianism; the strategy consists in modifying the Naïve
Principle, adding the proviso that any condition is fulfilled at some world:

Modal Meinongian Principle: For any condition on entities, there is a

unique entity satisfying exactly that condition at some world.

9781472524263_txt_print.indd 189 01/12/2015 09:44

 190 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

The Modal Meinongian Principle may superficially bear some resemblance

to Lewis’s principle of recombination (see Chapter 4), but it differs in key
respects. Unlike Lewis’s principle of recombination, the Modal Meinongian
Principle has no requirement for completeness or consistency of worlds: it is
as liberal as a principle can be with respect to conditions on entities.
While the debate on Meinongianism is traditionally kept separate from
the debate on alethic modalities, the remarks offered in this section should
suffice to show how closely interconnected the two debates are.

§8.5. Impossible worlds

Since the contributions of Morgan (1973) and Hintikka (1975), on one hand,
and Yagisawa (1988) and Mortensen (1989), on the other, which respec-
tively address logical and metaphysical issues regarding the impossible,
a distinctive stream of the discussion on modality has focused on impos-
sibilities and, especially, impossible worlds. Not only do these topics have
theoretical relevance per se; impossible worlds have been particularly handy
for explaining particular phenomena, such as the behavior of agents who hold
inconsistent beliefs, or the truth conditions of systems of propositions (e.g.
legal systems) that contain contradictory claims (see Lewis (1986) and Berto
(2013a)). In this section we shall introduce the key ideas and issues related to
impossibilities and impossible worlds.
The temptation to speak of impossible scenarios and impossible worlds
is pervasive; two lines of thought will help us appreciate this temptation. The
first combines theories of possible worlds with the basic sense of possibility.
In possible-worlds semantics, we are provided with a list of possible worlds;
we typically also assume that the list contains a certain number of worlds,
say N. Here, the sense of possibility pipes up: if N is the number of possible
worlds, surely we can think otherwise; there seems nothing contradictory in
thinking that this number could have been different. But we cannot represent
this possibility by means of a possible world, because we have agreed that N is
the number of possible worlds on the list and we are now considering additional
worlds to those on the list. Hence, we need to posit some impossible worlds in
order to account for the modal intuition that the number of possible worlds could
have been different. This line of reasoning is especially worrisome when applied
to realist theories of possible worlds and it was indeed initially entertained as a
reductio ad absurdum of Lewis’s modal realism (see Skyrms (1976) and Naylor
(1976)). Soon after, however, bolder theoretical proposals were put forward
which countenanced the existence of impossible worlds and exploited their
theoretical potential (see especially the articles in Priest (1997)).

9781472524263_txt_print.indd 190 01/12/2015 09:44

 NECESSITIES, NECESSARY EXISTENTS, AND THEIR BOUNDS 191

The second line of thought builds upon intuitions concerning impossible

scenarios and does not require the idea of possible worlds. In particular, it
rests on the hypothesis that some scenarios are not only counterfactuals
(against what actually is the case), but counterpossibles (against what
possibly is the case). If a certain contradictory claim were true, then what?
For instance, if Foffo were at the same time a cat and a dog, could Foffo eat
fish and potatoes? Counterpossibles enter the stage more frequently than it
might at first appear: we find them in constitutions and other crucial pieces
of legislation, we spot them in scientific papers, novels, movies, and—we
must confess—philosophy books. Since they have an impossible antecedent,
which turns out to be always false according to classical logic, counterpos-
sibles risk being trivially true; additionally, they cannot be analyzed in terms
of a possible-worlds semantics, because they involve scenarios that are not
caught by possible worlds. Yet, it would be desirable to be able to discriminate
between counterpossibles, sorting out those that rely upon intuitively sound
reasoning and those that do not. Positing the existence of impossible worlds
can help us do exactly that. See Brogaard and Salerno (2008) and Beall and
van Fraassen (2003) for some recent treatments of counterpossibles.
If we have good, intuitive reasons to consider admitting impossible
scenarios and impossible worlds to our modal theories, we also have some
good, intuitive reasons not to do so. As Berto (2013a) reminds us, the empir-
icist tradition maintained that we cannot entertain thoughts on the impossible
and that, while the practically impossible may perhaps be conceived (e.g. the
thought that Foffo can fly to the moon today), the theoretically impossible
is inconceivable (e.g. the thought that Foffo has round–square eyes). Lewis
(1986), for instance, rejects the idea that round-squares are conceivable as
well as the idea that there are impossible worlds (see also Vacek (2013)). Here
is a well-known passage in which Lewis (1986) illustrates why a realist about
worlds cannot accommodate the existence of impossible worlds:

suppose travellers told of a place in this world—a marvellous [sic]

mountain, far away in the bush—where contradictions are true. […] to tell
the alleged truth about the marvellously contradictory things that happen
on the mountain is no different from contradicting yourself. But there is
no subject matter, however marvellous, about which you can tell the truth
by contradicting yourself. Therefore there is no mountain where contradic-
tions are true. An impossible world where contradictions are true would be
no better. The alleged truth about its contradictory goings-on would itself
be contradictory. (7, fn. 3)

Since the end of the 1990s, a variety of strategies have been developed
to fix the language in which we talk about impossible worlds and, thus,

9781472524263_txt_print.indd 191 01/12/2015 09:44

 192 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

overcome objections, such as the one Lewis raises. The strategies exploit
the longstanding discussion of how to deal with apparent cases of contradic-
tions, also known as dialetheias. A dialetheia is a sentence such that both it
and its negation are true. Some apparent dialetheias, such as those arising
in connection with the so-called Liar’s Paradox, have been well-known for
millennia. Dialetheism is the view that accepts that there are dialetheias, and
tries to accommodate them within logic and language (see Priest and Berto
(2013) for an introductory discussion). The logic and language developed to
deal with dialetheism can be employed to provide a theory of impossible
worlds.
We now seem to have sufficient logical and conceptual resources to deal
with impossible scenarios and impossible worlds. Impossible scenarios and
worlds have the potential to considerably enrich the metaphysical appara-
tuses of the theories of possibility discussed in the previous chapters. Yet,
how innocent are impossible worlds from a metaphysical point of view?
What is the theoretical cost of accepting them into one’s metaphysics?
These questions remain to be fully addressed. We must ascertain, on a
case-by-case basis, whether and how each view (e.g. modal realism, modal
fictionalism, linguistic ersatzism, and so on) can integrate impossible worlds
into its theoretical framework. For some recent contributions on this issue,
see especially Yagisawa (2010) and Jago (2014).

9781472524263_txt_print.indd 192 01/12/2015 09:44

 NECESSITIES, NECESSARY EXISTENTS, AND THEIR BOUNDS 193

Study Questions
• What is the modal myopia that, according to Fine, characterizes the
contemporary debate about modalities?
• What is modal monism?
• What are some chief motivations for modal monism?
• What is modal pluralism?
• What are some chief motivations for modal pluralism?
• What is the distinction between worldly and transcendental necessary
truths?
• Put forward at least three different and original illustrations of
grounding.
• What is the relationship between grounding and alethic necessity?
• What is the relationship between grounding and non-alethic necessity?
• What is a necessary existent?
• What is Williamson’s alleged proof of necessary existence?
• What reasons can be advanced to resist Williamson’s proof?
• What is Meinongianism?
• What is naïve Meinongianism?
• What problems does naïve Meinongianism face?
• What strategies can be used to address the problems of naïve
Meinongianism?
• What are the two lines of reasoning that invite us to postulate
impossible worlds?
• What reasons do we have for not endorsing impossible worlds?
• What is a dialetheia?

FURTHER READING
Standard readings on the varieties of necessity are Sidelle (1989, 2002), as well
as Fine (2005). An essential reading, important for appreciating the contemporary
debate, is Kripke (1980); other classic readings include Farrell (1981), Hirsch (1986),
Plantinga (1992), Peacocke (1997), and Shoemaker (1998). For some recent work on
metaphysical necessity, see Drewery (2005), Williamson (2005), Bealer (2006), the
essays in Hale and Hoffman (2010), Cameron (2010), Nolan (2011), and Correia (2012).
Bliss and Trogdon (2014) and Raven (2015) provide an exhaustive and short
overview of metaphysical grounding, with an updated bibliography; other
starting points can be Clark and Liggins (2012) and Trogdon (2013). An opinionated
introductory reading on the topic is Fine (2012). In general, the essays contained in
Correia and Schnieder (2012) can be a good starting point to delve deeper into the
issue of metaphysical grounding.
▲

9781472524263_txt_print.indd 193 01/12/2015 09:44

 194 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

The argument for necessary existents is most concisely presented in

Williamson (2002). It builds upon previous work of Williamson—in particular his
works of 1990, 1998, and 2000—and the works of Linsky and Zalta (1994, 1996).
To appreciate the importance of the debate on necessary existents, see Prior and
Fine (1977: “Postscript”), Plantinga (1983), and Fine (1985). For recent discussion
of Williamson’s argument, see Rumfitt (2003), Wiggins (2003), Morato (2006),
Sider (2009), Stalnaker (2010), Efird (2010), Speaks (2012), Rasmussen (2013), and
Pruss and Rasmussen (2015).
A concise, and yet fairly exhaustive, initial reading on Meinongianism is Nelson
(2012). For some classic readings, the reader may want to start from Meinong’s
classic essay (1904) and then move to contemporary classics, including Parsons
(1979, 1980), Zalta (1983, 1988), and Priest (2005). The contemporary debate
on Meinongianism is lively, and often entangled with issues concerning alethic
modalities. For a recent study, see Berto (2012).
The main metaphysical issues surrounding impossible worlds are concisely
and clearly discussed in Berto (2013a) and Nolan (2013); for an introductory reading
on dialetheism, see Priest and Berto (2013); for the logical issues regarding
impossible worlds, see Beall and van Fraassen (2003), Priest (2005), and Goble
(2006). Classic readings on impossible worlds include Hintikka (1975), Yagisawa
(1988), and the essays contained in Priest (1997). For some recent and thorough
studies, see Yagisawa (2010) and Jago (2014). Other recent discussion pieces
include McDaniel (2004), Brogaard and Salerno (2008), Berto (2010, 2014), Jago
(2013a, 2013b), and Krakauer (2013).

9781472524263_txt_print.indd 194 01/12/2015 09:44

 Epilogue: Modal talk and the
analyses of modality

W e have now reached the conclusion of our analysis of the eight theories
of possibility that have emerged from the last fifty years of debate in
analytic metaphysics, and that now characterize this branch of philosophy.
Along the way, we encountered nineteen solutions to the PP, which are worth
recalling; let us list them:

EXP: The sentence in natural language: “It is possible that: s” expresses

a certain (conventionally codified) sentiment, on behalf of the person
who asserts it, toward s.
M: The sentence in natural language: “It is possible that: s” expresses the
existence of a modal fact of our world: the fact that it is possible that s.
MR*: The sentence of natural language: “It is possible that: s” expresses
the existence of worlds W1, …, Wn in which s*.
ME: Only one world exists, the actual world, of which possible worlds are
surrogates.
LE: The sentence in natural language: “It is possible that: s” expresses
a maximal and consistent description, given in a certain language,
according to which: s.
C: A state of affairs is possible if and only if it is obtained by means of a
legitimate recombination of constituents of at least one actual state
of affairs.
PEP: The sentence in natural language: “It is possible that: s” expresses
a proposition p, and is true if and only if there is a maximal and
consistent set of (complete and non-contradictory) propositions at
which p is true.
PES: The sentence in natural language: “It is possible that: s” expresses
a state of affairs S, and is true if and only if there is a maximal and
consistent set of states of affairs at which S is true.
PEI: The sentence in natural language: “It is possible that: s” expresses
an interpreted figure such that: the figure represents s; and,
according to the interpretation, the figure constitutes a maximal and
consistent representation.

9781472524263_txt_print.indd 195 01/12/2015 09:44

 196 A CRITICAL INTRODUCTION TO THE METAPHYSICS OF MODALITY

AE: The sentence in natural language: “It is possible that: s” expresses

an atomic entity S* that is abstract and maximal, and of which there
is at least one interpretation according to which S* counts as a repre-
sentation of the scenario expressed by the sentence “s”.
F: The sentence in natural language: “It is possible that: s” is true if and
only if its orthodox possible-worlds analysis is true.
RF: The sentence in natural language: “It is possible that: s” is true if and
only if, according to the story told in Lewis (1986), “s*” is true.
AF: The sentence in natural language: “It is possible that: s” is true if
and only if, according to the story told in Lewis (1986), there is a
collection of stories such that, according to at least one of them, s’s
equivalent “s*” is true.
F*: The sentence in natural language: “It is possible that: s” is true if and
only if, were the orthodox possible-worlds analysis to be true, then
the sentence would also be true.
F**: If we wish to judge in the most philosophically plausible way the
truth of the sentence in natural language: “It is possible that: s” then
we must assert that said sentence is true if and only if its orthodox
possible-worlds analysis is true.
AGN: The sentence in natural language: “It is possible that: s” is true if
and only if: (i) its equivalent “s*” in modal realism (or another chosen
possible-worlds theory) is true; and (ii) “s*” is admissible.
IE: The sentence in natural language: “It is possible that: s” expresses
the existence of an individual whose essence grounds the truth of
that sentence.
PE: The sentence in natural language: “It is possible that: s” expresses
the existence of a property whose essence grounds the truth of that
sentence.
D: The sentence in natural language: “It is possible that: s” expresses
the existence of one or more dispositions whose manifestation
includes the situation expressed by “s”.

To this already striking list, we should add the two skeptical options seen
in Chapter 2. Additionally, we should recall that, for each of the nineteen
solutions, we can ask how it would be affected by the varieties of necessity
and possibly enriched with Meinongianism or a theory of impossible worlds.
Two main lessons can be drawn from our lengthy discussion. The first
concerns the metaphysical status of modality: even if we endorse the most
powerful and radical theory—that is, modal realism—we are unable to fully
reduce modal entities to non-modal ones. Regardless of any reductive aspira-
tions, therefore, and given our best metaphysical theories, some modal
entities are nonetheless here to stay.

9781472524263_txt_print.indd 196 01/12/2015 09:44

 EPILOGUE: MODAL TALK AND THE ANALYSES OF MODALITY 197

The second lesson is that a piecemeal approach to the metaphysical

(and, arguably, philosophical) analysis of modality has the best payoffs. We
need not pick one view to fit all modal talk. We can choose, depending on
the sort of modal scenario in need of analysis. For instance, we can endorse
dispositionalism when it comes to natural necessity; a particular version of
fictionalism when it comes to mathematical necessity; linguistic ersatzism
when it comes to metaphysical necessity; and expressivism when it comes
to normative necessity. In particular, we should not think of possible-worlds
semantics as the only framework within which to discuss modality: we
should use such semantics where opportune, and drop it where some other
option seems to work better. This lesson is particularly in keeping with modal
pluralism, according to which different varieties of possibility and necessity
are autonomous.
The lessons in brief: no metaphysical theory of modality is reductive and
none is exhaustive. As for which theory should be chosen, if it is true that all
that which is is what it is partly in virtue of that which it could be, then we
can only hope that the theoretical possibilities we have offered in this book
are sufficiently many and diverse for a fruitful choice.
The pleasure of choosing, o reader, is now yours.

9781472524263_txt_print.indd 197 01/12/2015 09:44

9781472524263_txt_print.indd 198 01/12/2015 09:44
 Bibliography

Adams, R. M. (1974), “Theories of Actuality”. Noûs 8: 211–31.

Anderson, E. (2005), “How General is Generalized Scientific Essentialism?”
Synthese 144: 37–39.
Armour-Garb, B. (2015), “New Problems for Modal Fictionalism”. Philosophical
Studies 172: 1201–19.
Armstrong, D. M. (2004), “Theorie Combinatoire Revue et Corrigée.” In J. M.
Monnoyer (ed.), La Structure du Mond: Objets, Propriétés, États et Choses.
Paris: Vrin: 185–98.
Armstrong, D. M. (1997), A World of States of Affairs. Cambridge: Cambridge
University Press.
Armstrong, D. M. (1989), A Combinatorial Theory of Possibility. Cambridge:
Cambridge University Press.
Armstrong, D. M. (1986), “The Nature of Possibility”. Canadian Journal of
Philosophy 16: 575–94.
Armstrong, D. M. (1978), Nominalism and Realism (2 vol.). Cambridge:
Cambridge University Press.
Austin, C. J. (2015), “The Truthmaking Argument Against Dispositionalism.” Ratio
28: 271–85.
Ayer, A. (1936), Language, Truth, and Logic. London: Victor Gollancz Ltd.
Baldwin, T. (1998), “Modal Fictionalism and the Imagination”. Analysis 58: 72–5.
Ballarin, R. (2010), “Modern Origins of Modal Logic”. In E. N. Zalta (ed.), The
Stanford Encyclopedia of Philosophy (Winter 2014). Available from http://plato.
stanford.edu/archives/win2014/entries/logic-modal-origins/
Barrett, R. B. and R. F. Gibson (1990), Perspectives on Quine. Cambridge: Basil
Blackwell.
Bealer, G. (2006), “A Definition of Necessity”. Philosophical Perspectives 20:
17–39.
Bealer, G. (1987), “The Philosophical Limits of Scientific Essentialism”.
Philosophical Perspectives 1: 289–365.
Beall, J. C. and B. Van Fraassen (2003), Possibilities and Paradox. An Introduction
to Modal and Many-Valued Logic. Oxford: Oxford University Press.
Beebee, H. and N. Sabbarton-Leary (eds) (2010), The Semantics and
Metaphysics of Natural Kinds. New York: Routledge.
Bennett, J. (1994), “Descartes’s Theory of Modality”. The Philosophical Review
103: 639–67.
Berto, F. (2014), “On Conceiving the Inconsistent”. Proceedings of the
Aristotelian Society 114: 113–21.
Berto, F. (2013a), “Impossible Worlds”. In E. N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Winter 2013). Available from http://plato.stanford.
edu/archives/win2013/entries/impossible-worlds/

9781472524263_txt_print.indd 199 01/12/2015 09:44

 200 BIBLIOGRAPHY

Berto, F. (2013b), “Modal Meinongianism and Actuality”. Humana.Mente 25:

155–76.
Berto, F. (2012), Existence as a Real Property. The Ontology of Meinongianism.
Dordrecht: Synthèse Library, Springer.
Berto, F. (2010), “Impossible Worlds and Propositions: Against the Parity Thesis”.
The Philosophical Quarterly 60: 471–86.
Berto, F. (2007), How to Sell a Contradiction: The Logic and Metaphysics of
Inconsistency. London: College Publications.
Bird, A. (2009), “Essences and Natural Kinds”. In R. Le Poidevin (ed.), The
Routledge Companion to Metaphysics. New York: Routledge: 497–506.
Bird, A. (2007), Nature”s Metaphysics: Laws and Properties. Oxford: Oxford
University Press.
Bird, A. (2005), “Laws and Essences.” Ratio 18: 437–61.
Bird, A. (1998), “Dispositions and Antidotes”. Philosophical Quarterly 48: 227–34.
Bird, A. and E. Tobin (2015), “Natural Kinds”. In E. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Spring 2015). Available from http://plato.stanford.
edu/archives/spr2015/entries/natural-kinds/
Blackburn, S. (1998), Ruling Passions. Oxford: Clarendon Press.
Blackburn, S. (1993), Essays in Quasi-Realism. New York: Oxford University
Press.
Blackburn, S. (1984), Spreading the Word. New York: Oxford University Press.
Blackburn, S. (1971), “Moral Realism”. In J. Casey (ed.), Morality and Moral
Reasoning. London: Methuen: 101–24.
Bliss, R. and K. Trogdon (2014), “Metaphysical Grounding”. In E. Zalta (ed.), The
Stanford Encyclopedia of Philosophy (Winter 2014). Available from http://plato.
stanford.edu/archives/win2014/entries/grounding/
Blumenfeld, D. (1973), “Leibniz’s Theory of Striving Possibles”. Studia Leibnitiana
5: 163–77.
Bobzien, S. (2014), “Ancient Logic”. In E. N. Zalta (ed.), Stanford Encyclopedia of
Philosophy (Spring 2014). Available from http://plato.stanford.edu/archives/
spr2014/entries/logic-ancient/
Bobzien, S. (2011), “Dialectical School”. In E. N. Zalta (ed.), Stanford Encyclopedia
of Philosophy (Fall 2011). Available from http://plato.stanford.edu/archives/
fall2011/entries/dialectical-school/
Bobzien, S. (1998), Determinism and Freedom in Stoic Philosophy. Oxford:
Clarendon Press.
Boghossian, P. A. (2009), “Content”. In J. Kim and E. Sosa (eds), A Companion to
Metaphysics, 2nd edn. Oxford: Blackwell: 188–91.
Boghossian, P. A. and C. A. B. Peacocke (eds) (2000), New Essays on the A
Priori. Oxford: Oxford University Press.
Borghini, A. (2009), “Dispositions and Their Intentions”. In G. Damschen,
R. Schnepf, and K. Stueber (eds), Debating Dispositions: Issues in Metaphysics,
Epistemology, and Philosophy of Mind. Berlin: De Gruyter: 204–19.
Borghini, A. (2005), “Counterpart Theory and Mereological Essentialism”.
Dialectica 59: 67–74.
Borghini, A. (2003), “Un mondo di possibilità. Realismo modale senza mondi
possibili”. Rivista di Estetica 26: 87–100.
Borghini, A. and N. E. Williams (2008), “A Dispositional Theory of Possibility”.
Dialectica 62: 21–41.

9781472524263_txt_print.indd 200 01/12/2015 09:44

 BIBLIOGRAPHY 201

Bradley, R. (1989), “Possibility and Combinatorialism: Wittgenstein Versus

Armstrong”. Canadian Journal of Philosophy 19: 15–41.
Bricke, J. (2011), “Hume on Liberty and Necessity”. In E. S. Radcliffe (ed.),
A Companion to Hume, Oxford: Blackwell.
Bricker, P. (2007), “Concrete Possible Worlds”. In J. O’Leary-Hawthorne, T. Sider,
and D. Zimmerman (eds), Contemporary Debates in Metaphysics. Oxford:
Blackwell: 111–34.
Bricker, P. (2006a), “Absolute Actuality and the Plurality of Worlds”. Philosophical
Perspectives 20: 41–76.
Bricker, P. (2006b), “David Lewis: On the Plurality of Worlds”. In J. Shand (ed.),
Central Works of Philosophy, Vol. 5: The Twentieth Century: Quine and After.
London: Acumen Publishing: 246–67.
Brock, S. (1993), “Modal Fictionalism: A Response to Rosen”. Mind 102: 147–50.
Brogaard, B. (2006), “Two Modal-Isms: Fictionalism and Ersatzism”. Philosophical
Perspectives 20: 77–94.
Brogaard, B. and J. Salerno (2008), “Remarks on Counterpossibles”. Synthese
190: 639–60.
Bueno, O. and S. A. Shalkowski (2015), “Modalism and Theoretical Virtues:
Toward an Epistemology of Modality”. Philosophical Studies 172: 671–89.
Bueno, O. and S. A. Shalkowski (2013), “Logical Constants: A Modalist
Approach”. Noûs 47: 1–24.
Bueno, O. and S. A. Shalkowski (2009), “Modalism and Logical Pluralism”. Mind
118: 295–321.
Burgess, J. P. (1997), “Quinus ab Omni Naevo Vindicatus”. In A. A. Kazmi (ed.),
Meaning and Reference, Canadian Journal of Philosophy, Supp. Vol. 23:
25–65.
Cameron, R. P. (2010), “The Grounds of Necessity”. Philosophy Compass 5:
348–58.
Cameron, R. P. (2008), “Recombination and Intrinsicality”. Ratio 21: 1–12.
Campbell, J. K., M. O’Rourke, and M. Slater (eds) (2011), Carving Nature at
Its Joints. Natural Kinds in Metaphysics and Science. Cambridge, MA: MIT
Press.
Campbell, R. (1964), “Modality ‘De Dicto’; and ‘De Re’”. Australasian Journal of
Philosophy 42: 345–58.
Carey, S. (2009), The Origin of Concepts. New York: Oxford University Press.
Carnap, R. (1947), Meaning and Necessity. Chicago: University of Chicago Press.
Carnap, R. (1946), “Modalities and Quantification”. Journal of Symbolic Logic 11:
33–64.
Carnap, R. (1936), “Testability and Meaning I”. Philosophy of Science 3: 420–71.
Carrara, M., R. Ciuni, and G. Lando (2011), Composition, Counterfactuals, and
Causation. Special issue of Humana.Mente, 19.
Casullo, A. (2003), A Priori Justification. Oxford: Oxford University Press.
Chalmers, D. (2002), “Does Conceivability Entail Possibility?” In T. Gendler and
J. O’Leary-Hawthorne (eds), Conceivability and Possibility. Oxford: Oxford
University Press: 145–200.
Chase, J. and J. Reynolds (2010), Analytic Versus Continental: Arguments on the
Methods and Value of Philosophy. New York: Acumen.
Chellas, B. (1980), Modal Logic: An Introduction. Cambridge: Cambridge
University Press.

9781472524263_txt_print.indd 201 01/12/2015 09:44

 202 BIBLIOGRAPHY

Chignell, A. (2012), “Kant, Real Possibility, and the Threat of Spinoza”. Mind 121:
635–75.
Chihara, C. (1998), The Worlds of Possibility. Modal Realism and the Semantics
of Modal Logic. Oxford University Press, Oxford.
Clark, M. and D. Ligging (2012), “Recent Work on Grounding”. Analysis 72: 812–23.
Cohen, M. S. (2014), “Aristotle’s Metaphysics”. In E. N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Summer 2014). Available from <http://plato.
stanford.edu/archives/sum2014/entries/aristotle-metaphysics/
Contessa, G. (2013), “Dispositions and Interferences”. Philosophical Studies 165:
401–19.
Contessa, G. (2010), “Modal Truth-Makers and Two Varieties of Actualism”.
Synthese 174: 342–53.
Copeland, B. J. (2002), “The Genesis of Possibile Worlds Semantics”. Journal of
Philosophical Logic 31: 99–137.
Correia, F. (2012), “On the Reduction of Necessity to Essence”. Philosophy and
Phenomenological Research 84: 639–53.
Correia, F. (2010), “Grounding and Truth-Functions”. Logique et Analyse 53:
251–79.
Correia, F. and B. Schnieder (eds) (2012), Metaphysical Grounding:
Understanding the Structure of Reality. Cambridge: Cambridge University
Press.
Corry, R. (2011), “Can Dispositional Essences Ground the Laws of Nature?”
Australasian Journal of Philosophy 89: 263–75.
Craig, E. (2002), “The Idea of Necessary Connexion”. In P. Millican (ed.),
Reading Hume on Human Understanding. Oxford: Oxford University Press:
211–30.
Craig, W. L. (1988), The Problem of Divine Foreknowledge and Future
Contingents from Aristotle to Suárez. Leiden: Brill.
Cresswell, M. J. (1972), “The World Is Everything That Is the Case”. Australasian
Journal of Philosophy 50: 1–13.
Cunning, D. (2014), “Descartes’ Modal Metaphysics”. In E. N. Zalta (ed.), The
Stanford Encyclopedia of Philosophy (Spring 2014). Available from http://plato.
stanford.edu/archives/spr2014/entries/descartes-modal/
Curley, E. (1984), “Descartes on the Creation of the Eternal Truths”. The
Philosophical Review 93: 569–97.
Curley, E. and G. Walski (1999), “Spinoza’s Necessitarianism Reconsidered”. In R.
Gennaro and C. Huenemann (eds), New Essays on the Rationalists. Oxford:
Oxford University Press: 241–62.
Darby, G. and D. Watson (2010), “Lewis’s Principle of Recombination: Reply to
Efird and Stoneham”. Dialectica 64: 435–45.
Denby, D. A. (2008), “Generating Possibilities”. Philosophical Studies 141:
191–207.
Denby, D. A. (2006), “In Defense of Magical Ersatzism”. The Philosophical
Quarterly 56: 161–74.
Denyer, N. and S. Makin (2000), “Aristotle on Modality”. Proceedings of the
Aristotelian Society, Supplementary Volumes 74: 143–61, 163–78.
De Pierris, G. and M. Friedman (2013), “Kant and Hume on Causality.” In E. N.
Zalta (ed.), The Stanford Encyclopedia of Philosophy (Winter 2013). Available
from http://plato.stanford.edu/archives/win2013/entries/kant-hume-causality/>

9781472524263_txt_print.indd 202 01/12/2015 09:44

 BIBLIOGRAPHY 203

Dever, J. (2003), “Modal Fictionalism and Compositionality”. Philosophical

Studies 114: 223–51.
Divers, J. (2014a), “An Inconvenient Modal Truth”. Analysis 74: 575–7.
Divers, J. (2014b), “Mere Possibilities: Metaphysical Foundations for Modal
Semantics.” Philosophical Quarterly 64: 163–6.
Divers, J. (2014c), “The Modal Status of the Lewisian Analysis of Modality”. Mind
123: 861–72.
Divers, J. (2014d), “Modal Reality and (Modal) Logical Space”. Philosophy and
Phenomenological Research 88: 726–33.
Divers, J. (2013), “The Analysis of Possibility and the Extent of Possibility.”
Dialectica 67: 183–200.
Divers, J. (2012), “Modal Reality and (Modal) Logical Space”. Philosophy and
Phenomenological Research 88: 726–33.
Divers, J. (2007), “Quinean Skepticism about De Re Modality after David Lewis”.
European Journal of Philosophy 15: 40–62.
Divers, J. (2004), “Agnosticism about Other Worlds: A New Antirealist
Programme in Modality”. Philosophy and Phenomenological Research 69:
660–85.
Divers, J. (2002), Possible Worlds. London and New York: Routledge.
Divers, J. (1999), “A Modal Fictionalist Result”. Noûs 33: 317–46.
Divers, J. (1995), “Modal Fictionalism Cannot Deliver Possible Worlds
Semantics”. Analysis 55: 81–8.
Divers, J. and J. Hagen (2006), “The Modal Fictionalist’s Predicament”. In
F. McBride (ed.), Identity and Modality: New Essays in Metaphysics. Oxford:
Oxford University Press: 57–73.
Divers, J. and J. Melia (2003), “Genuine Modal Realism Limited.” Mind 112:
83–6.
Divers, J. and J. Melia (2002), “The Analytic Limit of Genuine Modal Realism.”
Mind 111: 15–36.
Dorr, C. (2005), “Propositions and Counterpart Theory.” Analysis 65: 210–18.
Drewery, A. (2005), “Essentialism and the Necessity of the Laws of Nature”.
Synthese 144: 381–96.
Dutihl Novaes, C. (2007), Formalizing Medieval Logical Theories: Suppositio,
Consequentiae and Obligationes. Dordrecht: Springer.
Dutihl Novaes, C. and S. Read (forthcoming), Cambridge Companion to Medieval
Logic.Cambridge: Cambridge University Press.
Eco, U. (1999), Kant and the Platypus: Essays on Language and Cognition.
Orlando, FL: Harcourt.
Efird, D. (2010), “Is Timothy Williamson a Necessary Existent?” In B. Hale and
A. Hoffmann (eds), Modality. Metaphysics, Logic, and Epistemology. Oxford:
Oxford University Press: 97–108.
Efird, D. and T. Stoneham (2006), “Combinatorialism and the Possibility of
Nothing.” Australasian Journal of Philosophy 84: 269–80.
Eklund, M. (2007), “Fictionalism”. In E. N. Zalta (ed.), The Stanford Encyclopedia
of Philosophy (Summer 2007). Available from http://plato.stanford.edu/
archives/sum2007/entries/fictionalism/
Ellis, B. (2002), The Philosophy of Nature—A Guide to the New Essentialism.
Montreal: McGill-Queens University Press.
Ellis, B. (2001), Scientific Essentialism. Cambridge: Cambridge University Press.

9781472524263_txt_print.indd 203 01/12/2015 09:44

 204 BIBLIOGRAPHY

Ellis, B. and C. Lierse (1994), “Dispositional Essentialism”. Australasian Journal of

Philosophy 72: 27–45.
Ereshefsky, M. (2010), “Microbiology and the Species Poblem”. Biology &
Philosophy 25: 553–68.
Fara, M. and T. Williamson (2005), “Counterparts and Actuality”. Mind 114: 1–30.
Farrell, R. (1981), “Metaphysical Necessity Is Not Logical Necessity”.
Philosophical Studies 39: 141–53.
Feldman, R. (2002), Epistemology. Upper Saddle River, NJ: Pearson.
Field, H. (1989), Realism, Mathematics and Modality. Oxford: Blackwell.
Field, H. (1980), Science Without Numbers. Oxford: Blackwell.
Fine, K. (2012), “A Guide to Ground”. In F. Correia and B. Schnieder (eds),
Metaphysical Grounding: Understanding the Structure of Reality. Cambridge:
Cambridge University Press: 37–80.
Fine, K. (2005), Modality and Tense: Philosophical Papers. Oxford: Oxford
University Press.
Fine, K. (2002), “The Varieties of Necessity”. In T. S. Gendler and O’Leary-
Hawthorne (eds), Conceivability and Possiblity. Oxford: Oxford University
Press: 253–81.
Fine, K. (2001), “The Question of Realism”. Philosopher’s Imprint 1: 1–30.
Fine, K. (1995), “The Logic of Essence”. Journal of Philosophical Logic 24:
241–73.
Fine, K. (1994), “Essence and Modality”. In J. Tomberlin (ed.), Philosophical
Perspectives 8: Logic and Language. Atascadero: Ridgeview Publishing: 1–16.
Fine, K. (1985), “Plantinga on the Reduction of Possibilist Discourse”. In J.
Tomberlin and P. Van Inwagen (eds), Alvin Plantinga. Dordrecht: D. Reidel:
145–86.
Fine, K. (1974), “Model Theory for Modal Logic Part I: the ‘De Re/De Dicto’
Distinction”. Journal of Philosophical Logic 7: 125–56.
Fitch, G. (1979), “Analyticity and Necessity in Leibniz”. Journal of the History of
Philosophy 17: 29–42.
Fitting, M. and R. L. Mendelsohn (1998), First-Order Modal Logic. Dordrecht:
Kluwer Academic Publishers.
Føllesdal, D. (1961), Referential Opacity and Modal Logic. Ph.D. Dissertation,
Harvard University, Cambridge, MA.
Forbes, G. (1992), “Melia on Modalism”. Philosophical Studies 68: 57–63.
Forbes, G. (1989), Languages of Possibility. An Essay in Philosophical Logic.
Oxford: Blackwell.
Forbes, G. (1985), The Metaphysics of Modality. Oxford: Oxford University Press.
Forbes, G. (1982), “Canonical Counterpart Theory”. Analysis 42: 33–7.
Friedman, R. L. and L. O. Nielsen (eds) (2003), The Medieval Heritage in Early
Modern Metaphysics and Modal Theory. Dordrecht: Kluwer.
Gabbay Dov, M. and J. Woods (eds) (2009), Handbook of the History of Logic
Volume 5. Logic from Russell to Church. Amsterdam: Elsevier.
Gabbay Dov, M. and J. Woods (eds) (2008a), Handbook of the History of Logic
Vol. 2. Medieval and Renaissance Logic. Amsterdam: Elsevier.
Gabbay Dov, M. and J. Woods (eds) (2008b), Handbook of the History of Logic
Vol. 4. British Logic in the Nineteenth Century. Amsterdam: Elsevier.
Gabbay Dov, M. and J. Woods (eds) (2006), Handbook of the History of Logic
Vol. 7. Logic and the Modalities in Twentieth Century. Amsterdam: Elsevier.

9781472524263_txt_print.indd 204 01/12/2015 09:44

 BIBLIOGRAPHY 205

Gabbay Dov, M. and J. Woods (eds) (2004a), Handbook of the History of Logic
Vol. 1. Greek, Indian and Arabic Logic. Amsterdam: Elsevier.
Gabbay Dov, M. and J. Woods (eds) (2004b), Handbook of the History of Logic
Vol. 3. The Rise of Modern Logic: Leibniz to Frege. Amsterdam: Elsevier.
Garrett, D. (2001), “Spinoza”s Necessitarianism”. In Y. Yovel (ed.), God and Nature
in Spinoza”s Metaphysics. Leiden: E. J. Brill: 191–218.
Garrett, D. (1985), “Priority and Separability in Hume’s Empiricism”. Archiv für
Geschichte der Philosophie 67: 270–88. Reprinted in D. W. D. Owen (ed.)
(2000), Hume: General Philosophy, The International Library of Critical Essays
in the History of Philosophy. Aldershot: Dartmouth Publishing Company: 71–89.
Garson, J. W. (2006), Modal Logic for Philosophers. Cambridge: Cambridge
University Press.
Gaskin, R. (1995), The Sea Battle and the Master Argument: Aristotle and
Diodorus Cronus on the Metaphysics of the Future. Berlin and New York: De
Gruyter.
Geirsson, H. (2014), “Conceivability and Coherence: A Skeptical View of
Zombies”. Erkenntnis 79: 211–25.
Geirsson, H. (2005), “Conceivability and Defeasible Modal Justification”.
Philosophical Studies 122: 279–304.
Gendler, T. S. and J. O’Leary-Hawthorne (eds) (2002), Conceivability and
Possibility. Oxford: Oxford University Press.
Gibbard, A. (2003), Thinking How to Live. Cambridge, MA: Harvard University
Press.
Gibbard, A. (1990), Wise Choices, Apt Feelings. Cambridge, MA: Harvard
University Press.
Girle (2003), Possible Worlds. Montreal: McGill–Queens University Press.
Girle (2000), Modal Logics and Philosophy. Teddington: Acumen.
Goble, L. (2006), “Paraconsistent Modal Logic”. Logique et Analyse 193: 3–29.
Haack, S. (1978), Philosophy of Logics. Cambridge: Cambridge University Press.
Hacking, I. (2009), The Social Construction of What? Cambridge, MA: Harvard
University Press.
Hale, B. (1995a), “A Desperate Fix”. Analysis 55: 74–81.
Hale, B. (1995b), “Modal Fictionalism: A Simple Dilemma”. Analysis 55: 63–7.
Hale, B. (1993), “Can There Be a Logic of Attitudes?”. In J. Haldane and C. Wright
(eds), Reality, Representation and Projection. Oxford: Oxford University
Press: 337–63.
Hale, B. and A. Hoffman (eds) (2010), Modality: Metaphysics, Logic, and
Epistemology. Oxford: Oxford University Press.
Harré, R. (1970), “Powers”. British Journal for the Philosophy of Science 21:
81–101.
Hasle, P. and P. Øhrstrøm (2011), “Future Contingents”. In E. N. Zalta (ed.), The
Stanford Encyclopedia of Philosophy (Summer 2011). Available from http://
plato.stanford.edu/archives/sum2011/entries/future–contingents/
Hawke, P. (2011), “Van Inwagen”s Modal Skepticism”. Philosophical Studies 153:
351–64.
Heil, J. (2003), From an Ontological Point of View. Oxford: Oxford University
Press.
Heller, M. (2005), “Anti–Essentialism and Counterpart Theory”. The Monist 88:
600–18.

9781472524263_txt_print.indd 205 01/12/2015 09:44

 206 BIBLIOGRAPHY

Heller, M. (2002), “Transworld Identity for the Ersatzist”. Philosophical Topics 30:
77–102.
Heller, M. (1998), “Property Counterparts in Ersatz Worlds”. The Journal of
Philosophy 30: 293–316.
Heller, M. (1996), “Ersatz Worlds and Ontological Disagreement”. Acta Analytica
15: 35–44.
Hendricks, V. F. and J. Symons (2005), Formal Philosophy. New York: Automatic
Press/VIP.
Hintikka, J. (1975), “Impossible Possible Worlds Vindicated”. Journal of
Philosophical Logic 4: 475–84.
Hintikka, J. (1973), Time and Necessity: Studies in Aristotle’s Theory of Modality.
Oxford: Oxford University Press.
Hintikka, J. (1963), “The Modes of Modality”. Acta Philosophica Fennica 16:
65–79.
Hintikka, J. (1961), “Modality and Quantification”. Theoria 27: 119–28.
Hintikka, J., U. Remes, and S. Knuuttila (1977), Aristotle on Modality and
Determinism. Amsterdam: North-Holland.
Hirsch, E. (2009), “Ontology and Alternative Languages”. In D. J. Chalmers,
D. Manley, and R. Wasserman (eds), Metametaphysics. New Essays on the
Foundations of Ontology. Oxford: Oxford University Press: 231–58.
Hirsch, E. (1997), Dividing Reality. Oxford: Oxford University Press.
Hirsch, E. (1986), “Metaphysical Necessity and Conceptual Truth”. Midwest
Studies in Philosophy 11: 243–56.
Huggett, N. (2010), “Zeno’s Paradoxes”. In E. N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Winter 2010). Available from http://plato.stanford.
edu/archives/win2010/entries/paradox-zeno/
Hughes, G. E. and M. J. Cresswell (1984), A Companion to Modal Logic.
London: Methuen.
Hughes, G. E. and M. J. Cresswell (1968), An Introduction to Modal Logic.
London: Methuen.
Humberstone, L. (2007), “Modal Logic for Other-World Agnostics: Neutrality and
Halldén Incompleteness”. Journal of Philosophical Logic 36: 1–32.
Hylton, P. (2007), Quine. New York: Routledge.
Ichikawa, J. and B. Jarvis (2012), “Rational Imagination and Modal Knowledge”.
Noûs 46: 127–58.
Jackson, F. and G. Priest (eds) (2004), Lewisian Themes: The Philosophy of David
K. Lewis. Oxford: Clarendon Press.
Jacobs, J. D. (2010), “A Powers Theory of Modality: Or, How I Learned to
Stop Worrying and Reject Possible Worlds”. Philosophical Studies 151:
227–48.
Jago, M. (2014), The Impossible. An Essay on Hyperintensionality. Oxford:
Oxford University Press.
Jago, M. (2013a), “Impossible Worlds”. Noûs 47: 713–28.
Jago, M. (2013b), “Against Yagisawa’s Modal Realism”. Analysis 73: 10–17.
Jauernig, A. (2008), “The Modal Strength of Leibniz’s Principle of the Identity of
Indiscernibles”. Oxford Studies in Early Modern Philosophy 4: 191–225.
Jenkins, C. S. (2010), “Concepts, Experience and Modal Knowledge”.
Philosophical Perspectives 24: 255–79.
Jubien, M. (2009), Possibility. Oxford: Oxford University Press.

9781472524263_txt_print.indd 206 01/12/2015 09:44

 BIBLIOGRAPHY 207

Jubien, M. (2007), “Analyzing Modality”. Oxford Studies in Metaphysics 3:

99–139.
Kail, P. (2003), “Conceivability and Modality in Hume. A Lemma in an Argument
in Defense of Skeptical Realism”. Hume Studies 29: 43–61.
Kalderon, M. (ed.) (2005), Fictionalism in Metaphysics. Oxford: Clarendon Press.
Kane, R. (ed.) (2002), The Oxford Handbook of Free Will. Oxford: Oxford
University Press.
Keskinen, A. (2012), “Quine on Objects and De Re Modality”. European Journal
of Analytic Philosophy 8: 4–17.
Keskinen, A. (2010), Quine’s Critique of Modal Logic and His Conception of
Objects. Tampere: Tampere University Press.
Khalidi, M. A. (2013), Natural Categories and Human Kinds. Classification in the
Natural and Social Sciences. Cambridge: Cambridge University Press.
Kim, S. (2005), “Modal Fictionalism and Analysis”. In M. Kalderon (ed.),
Fictionalism in Metaphysics. Oxford: Clarendon Press: 116–33.
Kim, S. (2002), “Modal Fictionalism Generalized and Defended”. Philosophical
Studies 111: 121–46.
Kim, S. (1986), “Possible Worlds and Armstrong’s Combinatorialism”. Canadian
Journal of Philosophy 16: 595–612.
Kistler, M. (2002), “The Causal Criterion of Reality and the Necessity of Laws of
Nature”. Metaphysica 3: 57–86.
Kitcher, P. (2000), “A Priori Knowledge Revisited”. In P. Boghossian and
C. Peacocke (eds), New Essays on the A Priori. Oxford: Oxford University
Press: 65–91.
Kitcher, P. (1980), “A Priori Knowledge”. The Philosophical Review 79: 3–23.
Klima, G. (2009), John Buridan. Oxford: Oxford University Press.
Kment, B. (2012), “Varieties of Modality”. In E. N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Winter 2012). Available from http://plato.stanford.
edu/archives/win2012/entries/modality-varieties/
Kment, B. (2006), “Counterfactuals and the Analysis of Necessity”. Philosophical
Perspectives 20: 237–302.
Kneale, W. and M. Kneale (1962), The Development of Logic. Oxford: Oxford
University Press.
Knuuttila, S. (2014), “Medieval Theories of Future Contingents”. In E. N. Zalta
(ed.), The Stanford Encyclopedia of Philosophy (Spring 2014). Available from
http://plato.stanford.edu/archives/spr2014/entries/medieval-futcont/
Knuuttila, S. (2013), “Medieval Theories of Modality”. In E. N. Zalta (ed.), The
Stanford Encyclopedia of Philosophy (Fall 2013). Available from http://plato.
stanford.edu/archives/fall2013/entries/modality-medieval/>
Knuuttila, S. (2008), “Medieval Modal Theories and Modal Logic”. In D. M.
Gabbay and J. Woods (eds), Handbook of the History of Logic Vol. 2. Medieval
and Renaissance Logic. Amsterdam: Elsevier: 505–78.
Knuuttila, S. (ed.) (1998), Modern Modalities. Studies of the History of Modal
Theories from Medieval Nominalism to Logical Positivism. Synthese Historical
Library 33: Dordrecht: Kluwer.
Knuuttila, S. and T. Kukkonen (2011), “Thought Experiment and Indirect Proof
in Averroes, Aquinas, and Buridan”. In K. Ierodiakonou and S. Roux (eds),
Thought Experiments In Methodological and Historical Perspectives. Leiden:
Brill: 83–99.

9781472524263_txt_print.indd 207 01/12/2015 09:44

 208 BIBLIOGRAPHY

Koistinen, O. (2003), “Spinoza’s Proof of Necessitarianism”. Philosophy and

Phenomenological Research 67: 283–310.
Krakauer, B. (2013), “What are Impossible Worlds?” Philosophical Studies 165:
989–1007.
Kripke, S. (1980), Naming and Necessity. Oxford: Blackwell.
Kripke, S. (1963), “Semantical Considerations on Modal Logics”. Acta
Philosophical Fennica 16: 83–94.
Kripke, S. (1959), “A Completeness Theorem in Modal Logics”. Journal of
Symbolic Logic 24: 1–14.
Kukkonen, T. (2005), “Possible Worlds in the Tahâfut al-falâsifa. Al-Gazâli on
Creation and Contingency”. Journal of the History of Philosophy 38: 479–502.
Kukkonen, T. (2000), “The Impossible insofar as it is Possible: Ibn Rushd and
Jean Buridan on Logic and Natural Theology”. In D. Perler and U. Rudolph
(eds), Logik und Theologie. Das Organon im arabischen und lateinischen
Mittelalter, Studien und Texte zur Geistesgeschichte des Mittelalters 84.
Leiden: Brill: 447–67.
Kung, P. (2011), “On the Possibility of Skeptical Scenarios”. European Journal of
Philosophy 19: 387–407.
Kung, P. (2010), “Imagining as a Guide to Possibility”. Philosophy and
Phenomenological Research 81: 620–63.
Lagerlund, H. (2000), Modal Syllogistics in the Middle Ages. Leiden: Brill.
Laporte, J. (2009), Natural Kinds and Conceptual Change. Cambridge:
Cambridge University Press.
Lewis, C. I. (1918), A Survey of Symbolic Logic. Berkeley, CA: University of
California Press. (Reprinted by Dover Publications: New York, 1960, with the
omission of Chapters 5 and 6.)
Lewis, C. I. and C. H. Langford (1932), Symbolic Logic. New York: Century
Company. (Reprinted by Dover Publications, New York, 1959, 2nd edn with an
Appendix III by C. I. Lewis.)
Lewis, F. A. (2009), “Parmenides’s Modal Fallacy”. Phronesis 54: 1–8.
Lewis, D. K. (1992), “Armstrong on Combinatorial Possibility”. Australasian
Journal of Philosophy 70: 211–224.
Lewis, D. K. (1991), Parts of Classes. Oxford: Basil Blackwell.
Lewis, D. K. (1986), On The Plurality of Worlds. Oxford: Clarendon Press.
Lewis, D. K. (1983), Philosophical Papers Vol. 1. Oxford: Clarendon Press.
Lewis, D. K. (1979), “Attitudes De Dicto and De se”. The Philosophical Review
88: 513–43.
Lewis, D. K. (1973), Counterfactuals. Oxford: Basil Blackwell.
Lewis, D. K. (1970), “Anselm and Actuality”. Noûs 4: 175–88.
Lewis, D. K. (1968), “Counterpart Theory and Quantified Modal Logic”. Journal of
Philosophy 65: 113–26.
Liggins, D. (2008), “Modal Fictionalism and Possible-Worlds Discourse”.
Philosophical Studies 138: 151–60.
Lin, M. (2012), “Rationalism and Necessitarianism”. Noûs 46: 418–48.
Linsky, B. (1991), “Is Lewis a Meinongian?” Australasian Journal of Philosophy
69: 438–53.
Linsky, B. (ed.) (1974), Reference and Modality. Oxford: Oxford University Press.
Linsky, B. and E. Zalta (1996), “In Defense of the Contingently Nonconcrete”.
Philosophical Studies 84: 283–94.

9781472524263_txt_print.indd 208 01/12/2015 09:44

 BIBLIOGRAPHY 209

Linsky, B. and E. Zalta (1994), “In Defense of the Simplest Quantified Modal
Logic”. In J. Tomberlin (ed.), Philosophical Perspectives 8: Logic and
Language. Atascadero: Ridgeview: 431–58.
Loewer, B. and J. Schaffer (2015), “A Companion to David Lewis”. Oxford:
Blackwell Publishers.
Look, B. C. (2013), “Leibniz’s Modal Metaphysics”. In E. N. Zalta (ed.), The
Stanford Encyclopedia of Philosophy (Spring 2013). Available from http://plato.
stanford.edu/archives/spr2013/entries/leibniz-modal/
Loux, M. J. (2006), Metaphysics: A Contemporary Introduction, 3rd edn. New
York: Routledge.
Loux, M. J. (ed.) (1979), The Possible and the Actual. Readings in the
Metaphysics of Modality. Ithaca, NY: Cornell University Press.
Lowe, E. J. (2012), “What is the Source of our Knowledge of Modal Truths?”
Mind 121: 919–50.
Lowe, E. J. (2006), The Four-Category Ontology: A Metaphysical Foundation for
Natural Science. Oxford: Oxford University Press.
Lycan, W. (1979), “The Trouble With Possibile Worlds”. In M. J. Loux (ed.), The
Possible and the Actual Readings in the Metaphysics of Modality. Ithaca, NY:
Cornell University Press: 274–316.
Machiavelli, N. (1513), The Prince. Trans. by H. C. Mansfield. Chicago: University
of Chicago Press, 1985.
Mackie, P. (2006), How Things Might Have Been: Individuals, Kinds, and
Essential Properties. Oxford: Oxford University Press.
Malink, M. (2006), “A Reconstruction of Aristotle’s Modal Syllogistic”. History and
Philosophy of Logic 27: 95–141.
Manley, D. and R. Wasserman (2011), “Dispositions, Conditionals, and
Counterexamples”. Mind 120: 1191–227.
Manley, D. and R. Wasserman (2008), “On Linking Dispositions and
Conditionals”. Mind 117: 59–84.
Marcus, R. B. (1993), Modalities: Philosophical Essays. Oxford: Oxford University
Press.
Marek, J. (2013), “Alexius Meinong”. In E. N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Fall 2013). Available from http://plato.stanford.
edu/archives/fall2013/entries/meinong/
Marmodoro, A. (2015), “Potentiality in Aristotle’s Metaphysics”. In K. Engelhard
and M. Quante (eds), The Handbook of Potentiality. Dordrecht: Springer:
forthcoming.
Martin, C. (2010), “A New Challenge to the Necessitarian Reading of Spinoza”.
Oxford Studies in Early Modern Philosophy 5: 25–70.
Martin, C. B. (1994), “Dispositions and Conditionals”. The Philosophical Quarterly
44: 1–8.
Martin, C. B. and J. Heil (1999), “The Ontological Turn”. Midwest Studies in
Philosophy 23: 34–60.
Mates, B. (1986), The Philosophy of Leibniz. Metaphysics and Language. Oxford:
Oxford University Press.
Mates, B. (1972), “Leibniz on Possible Worlds”. In H. Frankfurt (ed.), Leibniz. A
Collection of Critical Essays. New York: Doubleday: 335–64.
McDaniel, K. (2006), “Modal Realism”. Philosophical Perspectives: Metaphysics
20: 303–31.

9781472524263_txt_print.indd 209 01/12/2015 09:44

 210 BIBLIOGRAPHY

McDaniel, K. (2004), “Modal Realism with Overlap”. Australasian Journal of

Philosophy 82: 137–52.
McMichael, A. (1983), “A Problem for Actualism about Possible Worlds”. The
Philosophical Review 92: 49–66.
Meinong, A. (1904), “On Object Theory”. In R. Chisholm (ed.), Realism and the
Background of Phenomenology. Glencoe: The Free Press, 1960; translation
of “Über Gegenstandstheorie”. In A. Meinong (ed.), Untersuchungen zur
Gegenstandstheorie und Psychologie. Leipzig: Barth.
Melia, J. (2005), “Truthmaking Without Truthmakers”. In H. Beebee and J. Dodd
(eds), Truthmakers: The Contemporary Debate. Oxford: Oxford University
Press: 21–46.
Melia, J. (2003), Modality. Montreal and Kingston: McGill-Queen’s University
Press.
Melia, J. (2000), “Review of Charles S. Chihara, The Worlds of Possibility”. British
Journal for the Philosophy of Science 51: 333–7.
Melia, J. (1992), “Against Modalism”. Philosophical Studies 68: 35–56.
Mellor, D. H. (1974), “In Defense of Dispositions”. The Philosophical Review 83:
157–81.
Menzel, C. (2015), “Possible Worlds”. In E. N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Spring 2015). Available from http://plato.stanford.
edu/archives/spr2015/entries/possible-worlds/
Menzies, P. and P. Pettit (1994), “In Defence of Fictionalism about Possible
Worlds”. Analysis 54: 27–36.
Merricks, T. (2003), “The End of Counterpart Theory”. Journal of Philosophy 100:
521–49.
Merricks, T. (1999), “Composition as Identity, Mereological Essentialism, and
Counterpart Theory”. Australasian Journal of Philosophy 77: 192–5.
Mignucci, M. (1989), “Truth and Modality in Late Antiquity: Boethius on Future
Contingent Propositions”. In G. Corsi, C. Mangione and M. Mugnai (eds), Atti
del Convegno internazionale di storia della logica: le teorie delle modalità.
Bologna: CLUEB: 47–78.
Miller, J. A. (2001), “Spinoza’s Possibilities”. Review of Metaphysics 54: 779–815.
Molnar, G. (2003), Powers. Oxford: Oxford University Press.
Mondadori, F. (1985), “Understanding Superessentialism”. Studia Leibnitiana 17:
162–90.
Mondadori, F. (1983), “Counterpartese, Counterpartese*, CounterparteseD”.
Histoire, Epistémologie, Langage 5: 69–94.
Mondadori, F. and A. Morton (1976), “Modal Realism. The Poisoned Pawn”. The
Philosophical Review 85: 3–20.
Morato, V. (2006), “Propositions and Necessary Existents”. Grazer Philosophische
Studien 72: 211–31.
Morgan, C. G. (1973), “Systems of Modal Logic for Impossible Worlds” Inquiry
16: 280–9.
Mortensen, C. (1989), “Anything is Possible”. Erkenntnis 30: 319–37.
Mumford, S. (1998), Dispositions. Oxford: Oxford University Press.
Mumford, S. and R. Anjum (2011), Getting Causes from Powers. Oxford: Oxford
University Press.
Naylor, M. (1986), “A Note on David Lewis’s Realism about Possible Worlds”.
Analysis 46: 28–9.

9781472524263_txt_print.indd 210 01/12/2015 09:44

 BIBLIOGRAPHY 211

Neale, S. (2000), “On a Milestone of Empiricism”. In P. Kotatko and A. Orenstein

(eds), Knowledge, Language and Logic: Questions for Quine. Dordrecht:
Kluwer: 237–346.
Nelson, M. (2012), “Existence”. In E. N. Zalta (ed.), The Stanford Encyclopedia of
Philosophy (Winter 2012). Available from http://plato.stanford.edu/archives/
win2012/entries/existence/
Newlands, S. (2013), “Leibniz and the Ground of Possibility”. The Philosophical
Review 122: 155–87.
Nolan, D. (2015), “Noncausal Dispositions”. Noûs 49: 425–39.
Nolan, D. (2013), “Impossible Worlds”. Philosophy Compass 8: 360–72.
Nolan, D. (2011), “The Extent of Metaphysical Necessity”. Philosophical
Perspectives 25: 313–39.
Nolan, D. (2007), “Modal Fictionalism”. In E. N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Winter 2007). Available from http://plato.stanford.
edu/archives/win2007/entries/fictionalism–modal/
Nolan, D. (2005), David Lewis. Montreal and Kingston: McGill–Queen’s
University Press.
Nolan, D. (1997), “Three Problems for ‘Strong’ Modal Fictionalism”. Philosophical
Studies 87: 259–75.
Nolan, D. and J. O’Leary-Hawthorne (1996), “Reflexive Fictionalisms”. Analysis
56: 26–32.
Nolt, J. (1986), “What are Possible Worlds?”. Mind 95: 432–45.
Noonan, H. (1994), “In Defence of the Letter of Fictionalism”. Analysis 54: 133–9.
Normore, C. (1991), “Descartes’s Possibilities”. In G. J. D. Moyal (ed.), René
Descartes: Critical Assessments (Volume III). London and New York:
Routledge: 68–83.
Normore, C. (1982), “Future Contingents”. In N. Kretzmann, A. Kenny, and
J. Pinborg (eds), The Cambridge History of Later Medieval Philosophy.
Cambridge: Cambridge University Press: 358–81.
Palmer, J. (2012a), “Parmenides”. In E. N. Zalta (ed.), The Stanford Encyclopedia
of Philosophy (Summer 2012). Available from http://plato.stanford.edu/
archives/sum2012/entries/parmenides/
Palmer, J. (2012b), “Zeno of Elea”. In E. N. Zalta (ed.), The Stanford Encyclopedia
of Philosophy (Spring 2012). Available from http://plato.stanford.edu/archives/
spr2012/entries/zeno-elea/
Palmer, J. (2009), Parmenides and the Presocratic Philosophy, Oxford: Oxford
University Press.
Parsons, T. (1980), Nonexistent Objects. New Haven: Yale University Press.
Parsons, T. (1979), “The Methodology of Nonexistence”. The Journal of
Philosophy 76: 649–62.
Parsons, T. (1969), “Essentialism and Quantified Modal Logic”. The Philosophical
Review 78: 35–52.
Parsons, T. (1967), “Grades of Essentialism in Quantified Modal Logic”. Noûs 1:
181–91.
Patterson, R. (1995), Aristotle”s Modal Logic. Essence and Entailment in the
Organon. Cambridge: Cambridge University Press.
Peacocke, C. (1999), Being Known. Oxford: Oxford University Press.
Peacocke, C. (1997), “Metaphysical Necessity: Understanding, Truth,
Epistemology”. Mind 106: 521–74.

9781472524263_txt_print.indd 211 01/12/2015 09:44

 212 BIBLIOGRAPHY

Pike, N. (1977), “Divine Foreknowledge, Human Freedom, and Possible Worlds”.

The Philosophical Review 86: 209–16.
Plantinga, A. (2003), Essays in the Metaphysics of Modality. Oxford: Oxford
University Press.
Plantinga, A. (1992), The Nature of Necessity. Oxford: Clarendon Press.
Plantinga, A. (1985), “Reply to Kit Fine”. In J. Tomberlin and P. Van Inwagen (eds),
Alvin Plantinga: A Profile. Dordrecht: Reidel: 341–9.
Plantinga, A. (1983), “On Existentialism”. Philosophical Studies 44: 1–20.
Plantinga, A. (1974a), The Nature of Necessity. Oxford: Clarendon Press.
Plantinga, A. (1974b), God, Freedom, and Evil. New York: Harper & Row.
Popper, K. (1990), A World of Propensities. Bristol: Thoemmes.
Preyer, G. and F. Siebelt (eds) (2001), Reality and Humean Supervenience: Essays
on the Philosophy of David Lewis. Lanham, MD: Rowman and Littlefield.
Price, H. (2004), “Naturalism Without Representationalism”. In M. De Caro
and D. Macarthur (eds), Naturalism in Question. Cambridge, MA: Harvard
University Press: 71–88.
Priest, G. (2005), Towards Non-Being. The Logic and Metaphysics of
Intentionality. Oxford: Oxford University Press.
Priest, G. (ed.) (1997), “Impossible Worlds”. Special issue of the Notre Dame
Journal of Formal Logic 38: 481–660.
Priest, G. and F. Berto (2013), “Dialetheism”. In E. N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Summer 2013). Available from http://plato.
stanford.edu/archives/sum2013/entries/dialetheism/
Prior, A. N. (1957), Time and Modality. Oxford: Oxford University Press.
Prior, A. N. and K. Fine (1977), Worlds, Times and Selves. London: Duckworth.
Pruss, A. R. (2011), Actuality, Possibility, and Worlds. New York: Continuum.
Pruss, A. R. (2002), “The Actual and the Possible”. In R. M. Gale (ed.), Blackwell
Guide to Metaphysics. Oxford: Blackwell: 317–33.
Pruss, A. R. and J. Rasmussen (2015), Necessary Existence. Oxford: Oxford
University Press.
Quine, W. V. O. (1976), “Worlds Away”. The Journal of Philosophy 22: 859–63.
Quine, W. V. O. (1974), Roots of Reference. La Salle: Open Court.
Quine, W. V. O. (1953a), “Three Grades of Modal Involvement”. Proceedings
of the XIth International Congress of Philosophy, Vol. 11. Amsterdam: North-
Holland: 65–81.
Quine, W. V. O. (1953b), “Reference and Modality”. In W. V. O. Quine, From a
Logical Point of View. Cambridge, MA: Harvard University Press, 1980: 139–59.
Quine, W. V. O. (1947), “The Problem of Interpreting Modal Logic”. Journal of
Symbolic Logic 12: 43–8.
Rasmussen, J. (2013), “From Necessary Truth to Necessary Existence”. Polish
Journal of Philosophy 7: 19–30.
Raven, M. (2015), “Ground”. Philosophy Compass 10: 322–3.
Ray, G. (2000), “De Re Modality: Lessons from Quine”. In A. Orenstein A. and
P. Kotatko (eds), Knowledge, Language and Logic: Questions for Quine,
Boston Studies in the History and Philosophy of Science, Vol. 210. Dordrecht:
Kluwer Academic: 347–67.
Read, S. (forthcoming), “Logic in the XIVth Century”. In C. Dutilh Novaes and S.
Read (eds), Cambridge Companion to Medieval Logic. Cambridge: Cambridge
University Press.

9781472524263_txt_print.indd 212 01/12/2015 09:44

 BIBLIOGRAPHY 213

Rescher, N., R. Manor, A. Vander Nat, and Z. Parks (eds) (1974), Studies in
Modality. Oxford: Blackwell.
Rini, A. (2010), Aristotle’s Modal Proofs: Prior Analytics A 8-22 in Predicate Logic,
The New Synthese Historical Library 68. Dordrecht: Springer.
Robertson, T. and P. Atkins (2013), “Essential vs. Accidental Properties”.
In E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Winter
2013). Available from http://plato.stanford.edu/archives/win2013/entries/
essential–accidental/
Roca–Royes, S. (2011), “Conceivability and De Re Modal Knowledge”. Noûs 45:
22–49.
Rorabiji, R. (1980), Necessity, Cause and Blame. Perspectives on Aristotle’s
Theory. Ithaca, NY: Cornell University Press.
Rosen, G. (1995), “Modal Fictionalism Fixed”. Analysis 55: 67–73.
Rosen, G. (1993), “A Problem for Fictionalism about Possible Worlds”. Analysis
53: 71–81.
Rosen, G. (1990), “Modal Fictionalism”. Mind 99: 327–54.
Routley, R. (1980), Exploring Meinong’s Jungle and Beyond. Canberra: Australian
National University Press.
Roy, T. (1995), “In Defense of Linguistic Ersatzism”. Philosophical Studies 80:
217–42.
Rudder Beker, L. (1985), “Was Leibniz Entitled to Possible Worlds?” Canadian
Journal of Philosophy 15: 57–74.
Rumfitt, I. (2003), “Contingent Existents”. Philosophy 78: 461–81.
Russell, B. (2014), “A Priori Justification and Knowledge”. In E. N. Zalta (ed.) The
Stanford Encyclopedia of Philosophy (Summer 2014). Available from http://
plato.stanford.edu/archives/sum2014/entries/apriori/
Ryle, G. (1949), The Concept of Mind. London: Hutchinson.
Sauchelli, A. (2013), “Modal Fictionalism, Possible Worlds, and Artificiality”. Acta
Analytica 28: 411–21.
Schrenk, M. (2010), “The Powerlessness of Necessity”. Noûs 44: 725–39.
Schrenk, M. (2009), “Hic Rhodos, Hic Salta: From Reductionist Semantics to
a Realist Ontology of Forceful Dispositions”. In G. Damschen, R. Schnepf,
and K. Stueber (eds), Debating Dispositions: Issues in Metaphysics,
Epistemology, and Philosophy of Mind. Berlin: De Gruyter: 145–67.
Shalkowski, S. (1994), “The Ontological Ground of the Alethic Modality”. The
Philosophical Review 103: 669–88.
Shoemaker, S. (1998), “Causal and Metaphysical Necessity”. Pacific Philosophical
Quarterly 79: 59–77.
Shoemaker, S. (1980), “Causality and Properties”. In P. Van Inwagen (ed.),
Time and Cause: Essays Presented to Richard Taylor. Dordrecht: D. Reidel
Publishing: 109–36.
Shroeder, M. (2010), Noncognitivism in Ethics. London: Routledge.
Shroeder, M. (2008a), “Expression For Expressivists”. Philosophy and
Phenomenological Research 76: 86–116.
Shroeder, M. (2008b), “How Expressivists Can and Should Solve their Problem
with Negation”. Noûs 42: 573–99.
Sidelle, A. (2009), “Conventionalism and the Contingency of Conventions”. Noûs
43: 224–41.
Sidelle, A. (2002), “On the Metaphysical Contingency of Laws of Nature”. In

9781472524263_txt_print.indd 213 01/12/2015 09:44

 214 BIBLIOGRAPHY

T. Gendler and J. O’Leary-Hawthorne (eds), Conceivability and Possibility.

Oxford: Oxford University Press: 309–36.
Sidelle, A. (1989), Necessity, Essence, and Individuation. Ithaca, NY: Cornell
University Press.
Sider, T. (2009), “Williamson’s Many Necessary Existents”. Analysis 69: 250–8.
Sider, T. (2006), “Beyond the Humphrey Objection”, ms. Available from http://
tedsider.org/papers/counterpart_theory.pdf.
Sider, T. (2005), “Another Look at Armstrong’s Combinatorialism”. Noûs 39: 680–96.
Sider, T. (2003), “Reductive Theories of Modality”. In M. J. Loux and
D. Zimmerman (eds), The Oxford Handbook of Metaphysics. Oxford: Oxford
University Press: 180–208.
Sider, T. (2002), “The Ersatz Pluriverse”. The Journal of Philosophy 99: 279–315.
Sider, T. (2001), “Review of Charles S. Chihara, The Worlds of Possibility”. The
Philosophical Review 110: 88–91.
Siderits, M. and J. D. O’Brien (1976), “Zeno and on Nāgārjuna on Motion”.
Philosophy East and West 26: 281–99.
Skyrms, B. (1981), “Tractarian Nominalism”. Philosophical Studies 40: 199–206.
Skyrms, B. (1976), “Impossible Worlds, Physics and Metaphysics”. Philosophical
Studies 30: 323–32.
Smith, R. (2014), “Aristotle’s Logic”. In E. N. Zalta, The Stanford Encyclopedia of
Philosophy (Spring 2014). Available from http://plato.stanford.edu/archives/
spr2014/entries/aristotle-logic/
Smullyan, A. F. (1948), “Modality and Description”. Journal of Symbolic Logic 13:
31–7.
Soames, S. (2010), Philosophy of Language. Princeton, NJ: Princeton University
Press.
Sosa, E. (2000), “Modal and Other A Priori Epistemology: How Can We
Know What is Possible and What is Impossible?”. The Southern Journal of
Philosophy XXXVIII (Supplement): 1–16.
Spade, P. V. (ed.) (1999), The Cambridge Companion to Ockham. Cambridge:
Cambridge University Press.
Spade, P. V. and C. Panaccio (2011), “William of Ockham”. In E. N. Zalta (ed.), The
Stanford Encyclopedia of Philosophy (Fall 2011). Available from http://plato.
stanford.edu/archives/fall2011/entries/ockham/
Speaks, J. (2012), “On Possibly Nonexistent Propositions”. Philosophy and
Phenomenological Research 85: 528–62.
Stalnaker, R. (2012), Mere Possibilities. Metaphysical Foundations of Modal
Metaphysics. Princeton, NJ: Princeton University Press.
Stalnaker, R. (2010), “Merely Possible Propositions”. In B. Hale and A. Hoffman
eds, Modality: Metaphysics, Logic, and Epistemology. Oxford: Oxford
University Press: 21–32.
Stalnaker, R. (2003), Ways a World Might Be. Metaphysical and
Anti-Metaphysical Essays. Oxford: Oxford University Press.
Stalnaker, R. (1996), “Impossibilities”. Philosophical Topics 24: 193–204.
Stalnaker, R. (1984), Inquiry. Cambridge, MA: MIT Press.
Strawson, G. (2000), “David Hume: Objects and Power”. In R. Read and K. A.
Richman (eds), The New Hume Debate. New York: Routledge: 31–51.
Strawson, G. (1989), The Secret Connexion: Causation, Realism and David
Hume. Oxford: Clarendon Press.

9781472524263_txt_print.indd 214 01/12/2015 09:44

 BIBLIOGRAPHY 215

Swoyer, C. (1982), “The Nature of Natural Laws”. Australasian Journal of

Philosophy 60: 203–23.
Thom, P. (2003), Medieval Modal Systems. Problems and Concepts. Ashgate
Studies in Medieval Philosophy, Aldershot Ashgate.
Thomas, H. G. (1996), “Combinatorialism and Primitive Modality”. Philosophical
Studies 83: 231–22.
Thomasson, A. (2010), “Modal Expressivism and the Methods of Metaphysics”.
Philosophical Topics 35: 135–60.
Torza, A. (2015), “Speaking of Essence”. The Philosophical Quarterly 65: 211–31.
Torza, A. (2011), “Models for Counterparts”, Axiomathes 21: 553–79.
Trogdon, K. (2013), “An Introduction to Grounding”. In M. Hoeltje, B. Schnieder,
and A. Steinberg (eds), Varieties of Dependence. Ontological Dependence,
Grounding, Supervenience, Response-Dependence. München: Philosophia:
97–122.
Vacek, M. (2013), “Concrete Impossible Worlds”. Filozofia 68: 523–29.
Van Fraassen, B. C. (1980), The Scientific Image. Oxford: Clarendon Press.
Van Fraassen, B. C. (1977), “The Only Necessity is Verbal Necessity”. Journal of
Philosophy 74: 71–85.
Van Inwagen, P. (1998), “Modal Epistemology”. Philosophical Studies 92: 67–84.
Van Inwagen, P. (1986), “Two Concepts of Possible Worlds”. In P. A. French, T. E.
Uehling, and H. K. Wettstein, Jr. (eds), Midwest Studies in Philosophy, XI:
Studies in Essentialism. Minneapolis: University of Minnesota Press: 185–213.
Van Inwagen, P. (1985), “Plantinga on Trans-World Identity”. In J. Tomberlin and
P. Van Inwagen (eds), Alvin Plantinga: A Profile. Dordrecht: Reidel: 101–20.
Van Inwagen, P. (1983), An Essay on Free Will. Oxford: Clarendon Press.
Van Inwagen, P. (1980), “Indexicality and Actuality”. The Philosophical Review 89:
403–26.
Van Rijen, J. (1989), Aspects of Aristotle’s Logic of Modalities. Synthese
Historical Library 35. Dordrecht: Kluwer.
Van Roojen, M. (2014), “Moral Cognitivism vs. Non-Cognitivism”. Stanford
Encyclopedia of Philosophy (Fall 2014). Available from http://plato.stanford.
edu/archives/fall2014/entries/moral–cognitivism/
Vance, C. (2014), “Dispositional Modal Truthmakers and the Necessary Origin”.
Philosophia 42: 1111–27.
Varzi, A. C. (2002), “Words and Objects”. In A. Bottani, M. Carrara, and D.
Giaretta (eds), Individuals, Essence, and Identity. Themes of Analytic
Metaphysics. Dordrecht: Kluwer: 49–75.
Varzi, A. C. (2001), “Parts, Counterparts, and Modal Occurrents”. Travaux de
Logique 14: 151–71.
Vetter, B. (2015), Potentiality. From Dispositions to Modality. Oxford: Oxford
University Press.
Vetter, B. (2014), “Dispositions without Conditionals”. Mind 123: 129–56.
Vetter, B. (2013a). “<<Can>>“ without Possible Worlds: Semantics for
Anti-Humeans”. Philosopher’s Imprint 13: 1–27.
Vetter, B. (2013b), “Multi-track Dispositions”. Philosophical Quarterly 63: 330–52.
Vetter, B. (2011a), “Recent Work: Modality without Possible Worlds”. Analysis 71:
742–54.
Vetter, B. (2011b), “On Linking Dispositions and Which Conditionals?”. Mind 120:
1173–89.

9781472524263_txt_print.indd 215 01/12/2015 09:44

 216 BIBLIOGRAPHY

Vihvelin, K. (1988), “The Modal Argument for Incompatibilism”. Philosophical

Studies 53: 227–44.
Von Wright, G. (1951), An Essay in Modal Logic. Amsterdam: North-Holland.
Wang, J. (2015), “The Modal Limits of Dispositionalism”. Noûs 49: 454–69.
Wang, J. (2013), “From Combinatorialism to Primitivism”. Australasian Journal of
Philosophy 91: 535–54.
Waterloo, S. (1982), Passage and Possibility. A Study of Aristotle’s Modal
Concepts. Oxford: Oxford University Press.
Weatherson, B. and D. Marshall (2012), “Intrinsic vs. Extrinsic Properties”. In
E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Fall 2014). Available
from http://plato.stanford.edu/archives/fall2014/entries/intrinsic-extrinsic/
Wee, C. (2006), “Descartes and Leibniz on Human Free Will and the Human
Ability to Do Otherwise”. Canadian Journal of Philosophy 36: 387–414.
Wiggins, D. (2003), “Existence and Contingency: A Note”. Philosophy 78:
483–94.
Wiggins, D. (1980), Sameness and Substance. Oxford: Basil Blackwell.
Williamson, T. (2013), Modal Logic as Metaphysics. Oxford: Oxford University
Press.
Williamson, T. (2010), “Modal Logic within Counterfactual Logic”. In B. Hale and
A. Hoffman (eds), Modality: Metaphysics, Logic, and Epistemology. Oxford:
Oxford University Press: 81–96.
Williamson, T. (2007a), The Philosophy of Philosophy. Oxford: Basil Blackwell.
Williamson, T. (2007b), “Philosophical Knowledge and Knowledge of
Counterfactuals”. Grazer Philosophische Studien 74: 89–123.
Williamson, T. (2005), “Armchair Philosophy: Metaphysical Modality and
Counterfactual Thinking”. Proceedings of the Aristotelian Society 105: 1–23.
Williamson, T. (2002), “Necessary Existents”. In A. O’Hear (ed.), Royal Institute of
Philosophy Supplement. Cambridge: Cambridge University Press: 269–87.
Williamson, T. (2000), “The Necessary Framework of Objects”. Topoi 19: 201–8.
Williamson, T. (1998), “Bare Possibilia”. Erkenntnis 48: 257–73.
Williamson, T. (1990), “Necessary Identity and Necessary Existence”. In R. Haller
and J. Brandl (eds), Wittgenstein—Towards a Re-evaluation. Proceedings of
the 14th International Wittgenstein Symposium, Vol. 1. Vienna: Hoder-Pichler-
Tempsky: 168–75.
Wilson, C. (2000), “Plenitude and Compossibility in Leibniz”. The Leibniz Review
10: 1–20.
Wittgenstein, L. (1953), Philosophical Investigations. Oxford: Basil Blackwell.
Wittgenstein, L. (1921), “Logisch–philosophische Abhandlung”. In Annalen der
Naturphilosophie 14: 185–262; revised English edition, Tractatus Logico-
Philosophicus. London: Kegan Paul, Trench, Trubner & Co.
Woodward, R. (2012), “Fictionalism and Incompleteness”. Noûs 46: 781–90.
Woodward, R. (2011), “Is Modal Fictionalism Artificial?“ Pacific Philosophical
Quarterly 92: 535–50.
Woodward, R. (2008), “Why Modal Fictionalism is Not Self-Defeating”.
Philosophical Studies 139: 273–88.
Wright, C. (1980), Wittgenstein on the Foundations of Mathematics. London:
Duckworth.
Yablo, S. (2001), “Go Figure: A Path Through Fictionalism”. Midwest Studies in
Philosophy 25: 72–102.

9781472524263_txt_print.indd 216 01/12/2015 09:44

 BIBLIOGRAPHY 217

Yablo, S. (1994), “Is Conceivability a Guide to Possibility?”. Philosophy and

Phenomenological Research 53: 1–42.
Yagisawa, T. (2010), Worlds and Individuals. Possible and Otherwise. Oxford:
Oxford University Press.
Yagisawa, T. (1988), “Beyond Possible Worlds”. Philosophical Studies 53:
175–204.
Yates, D. (2015), “Dispositionalism and the Modal Operators.” Philosophy and
Phenomenological Research 91: 411–24.
Zalta, E. N., (2010), “Logic and Metaphysics”. Journal of Indian Council of
Philosophical Research 27: 155–84.
Zalta, E. N. (2006), “Essence and Modality”. Mind 115: 659–93.
Zalta, E. N. (1988), Intentional Logic and the Metaphysics of Intentionality.
Cambridge, MA: MIT Press.
Zalta, E. N. (1983), Abstract Objects. An Introduction to Axiomatic Metaphysics.
Cambridge, MA: MIT Press.
Zupko, J. (2014), “John Buridan”. In E. N. Zalta (ed.), The Stanford Encyclopedia
of Philosophy (Spring 2014). Available from http://plato.stanford.edu/archives/
spr2014/entries/buridan/

9781472524263_txt_print.indd 217 01/12/2015 09:44

9781472524263_txt_print.indd 218 01/12/2015 09:44
 Index of Names

Abelard 35 Fine, Kit 15, 76, 82, 83, 159, 160–4,

Adams, Robert 130 173, 177–85, 193
Aenesidemus 49 Forbes, Graeme 61, 76, 77, 80, 81,
Al-Fārābī 35 83, 159–64, 173, 177–84, 185,
Al-Ghazali 35 193
Arcesilaus 49, 50
Aristotle 8, 15, 20, 24, 25, 28, 29, 30, Gersonides 35
33, 34, 35, 48, 75, 165
Armstrong, David 9, 78, 100, 116, Hale, Bob 145
122–30, 140, 142 Hegel, Georg W. F. 46
Athens 24, 25, 62 Hintikka, Jaakko 86, 118
Averroes 35 Hume, David 27, 39, 42, 45–6, 48, 49,
Avicenna 35 64, 98, 165

Babylonians 20 Inwagen, Peter van 50, 57–9, 71, 114,

Bird, Alexander 163 130
Boethius see Severinus Boethius
Boole, George 46 Jubien, Michael 160, 163, 173
Brock, Stuart 144
Bueno, Octavio 76, 85 Kant, Immanuel 27, 45
Buridan, John 36, 38–9, 48 Kātibī 35
Kilwarbi 35
Carnap, Rudolf 14, 86, 116, 118, 130, Kripke, Saul 57, 83, 86, 93, 173, 178,
167 179, 193
Carneades 49
Chihara, Charles 76, 80–5 Leibniz, Gottfried W. 8, 14, 39, 42–4,
Chrysippus 34 45, 48, 116, 130
Cretans 20 Lewis, Clarence I. 46, 51, 54
Lewis, David K. 9, 13, 14, 33, 45, 57,
Descartes, René 39, 40–1, 48, 49 59, 80, 87, 88, 91–114, 121,
Diodorus Cronus 31, 33, 35 123, 126, 131, 135, 147, 148,
Diogenes Laertius 21 149, 168, 184, 187, 188–92
Divers, John 7, 15, 57, 103, 140, Locke, John 83
149–54 Lowe, Jonathan 76, 82–5, 90, 106,
160–4, 173, 178
Egyptians 20
Ellis, Brian 160, 163 Machiavelli, Niccolò, 61
Epicurus 14 Megara see Megarian school (Index
Euclides of Megara 31 of Terms)

9781472524263_txt_print.indd 219 01/12/2015 09:44

 220 INDEX OF NAMES

Meinong, Alexius 187 see also Ryle, Gilbert 165

Meinongianism (Index of Terms)
Melia, Joseph 77, 89 Scotus, Duns 35, 45
Michel de Montaigne, 49 Severinus Boethius 31, 34, 35
Mill, John S. 46 Sextus Empiricus 49
Moore, George E. 49 Shalkowski, Scott 76, 85
Musil, Robert 1, 19, 92 Sider, Ted 82
Skirms, Brian 100
Ockham, William of 36–9, 48 Socrates 24
Spinoza, Baruch, 21, 39, 41–2, 45,
Parmenides 20, 21, 24, 41, 42, 48 48, 64
Philo 31, 33, 35, 45 Stagira 2
Plantinga, Alvin 80, 116, 130, 132, Stalnaker, Robert 116, 130
160, 161, 185 Swift, Jonathan 120
Plato 24, 25, 50
Popper, Karl 166 Thomasson, Amie 67
Price, Huw 60
Prior, Arthur 76 Williamson, Timothy 134, 160, 176,
Putnam, Hilary 49 185–7, 194
Pythagora 21 Wittgenstein, Ludwig 14, 49, 86, 116,
125, 130
Quine, Willard van Orman 7, 51–7, 59,
65, 70, 71, 167, 177 Yablo, Stephen 59

Rosen, Gideon 14, 80, 81, 140–4, 155 Zeno of Elea 20–5, 41, 42, 47

9781472524263_txt_print.indd 220 01/12/2015 09:44

 Index of Terms

a priori/a posteriori 16, 63, 71, 85 composition as identity 96 see also

absolute necessity 26 see also mereology
conditional necessity conceivability 46
abstract/concrete concepts 10, 12
entity 16 concrete world 95 see also abstract/
possible world see concrete world concrete
academic skepticism see skepticism concretism 91 see also modal
accessibility relation 87 realism
accidental 8 conditional necessity 26 see also
acquaintance see knowledge absolute necessity
actuality 5 conditionals 25, 30, 31–3
priority of actuality 117 consistency 119, 132
admissible sentence 152 constituent see state of affair
advanced modalizing 102, 110 contraction see possible world
adverbs, modalities as 11 counterfactual 9, 33, 88, 101, 107,
agency 9 144–5, 160, 191
agnosticism see modal agnosticism counterpart theory 87, 93, 98, 99,
alethic modalities 9, 10, 11, 12, 13, 100–12, 113, 116, 117, 126, 146,
16, 175, 176 153, 159, 164
alien possibility 105–9, 128–30, 132, counterpossible 191 see also
133, 137, 154, 171, 179 counterfactual
analytic hypothesis see philosophical
method de re/de dicto 50, 51–7, 66, 70
analytic philosophy 6 deduction 5
arabic philosophy on modality 35–6 deontic modalities 10
attributes 42 determinism 42
diachronic possibility 25, 28
bivalence, principle of 30 dialetheia 192 see also dialetheism;
impossible worlds
capacities 25, 28, 51 see also dialetheism 11, 177, 186, 192, 194 see
dispositions; powers also impossible worlds
cardinality see possible worlds dispositionalism 15, 28–9, 164–72,
cause 25, 97 173
ceteris paribus 166 dispositions 28–9, 51, 164, 165–8,
combinatorialism see ersatzism 173 see also capacities; powers
complete concept 43 degree 170
completeness 119 see also divided modality 37–9
maximality divine foreknowledge 34
composite modality 37–9 duplication see intrinsic duplicate

9781472524263_txt_print.indd 221 01/12/2015 09:44

 222 INDEX OF TERMS

early modern philosophy on modality future contingents 9, 18, 25, 30–1,

39–46 48, 97
element 25
enunciative operators 11 God’s omnipotence see omnipotence
epistemic modalities 10 grounding 15, 176, 182, 192
epistemic problem of possibility see grounds see grounding
problem of possibility
epistemology 8 history of modality 18, 47–8
ersatzism 14, 115–37, 138 humean recombination see
atomic 116, 135–7, 138 recombination principle
combinatorialism 116, 122–30, 138 humility 108
see also states of affairs Humphrey objection 102
ersatz 116
linguistic 44, 116, 118–22 images 134–5 see also ersatzism,
magical see ersatzism, atomic pictorial
pictorialism 116, 130–5, 138 imagination and possibility 58–9, 103
surrogate 117 impossibility 9, 16, 40–1, 129, 162
essence 8, 15, 51, 82, 160–4, 176 see scenarios 3, 21, 40–1
also essentialism worlds 16, 18, 110, 111, 119, 159,
individual 133, 159 190–2, 194, 196
essentialism 15, 18, 25, 36 see also independence of accompaniment 99,
grounding 111
Aristotelian essentialism 54 indian philosophy 20
ethics 8 individual 96, 123–30 see also
individual 160–4, 173 essence
intrinsic 163, 173 induction 5
modalism and 82–5 instantiation 39, 123–4, 125 see also
new essentialism 15 property
property 160–4, 173 interpolation see property
scientific 163, 173 interpretation, semantic see semantic
super-essentialism 42, 44 interpretation
event 96 intrinsicness
excluded middle, principle of 30 duplicate 99, 109–12
exemplification 189 see also property property 99
expressivism see modal see also essentialism
expressivism isolatedness of worlds 95, 96
extension 42
external relation 110 see also Jewish philosophy 20
maximality
extraction see possible world knowledge
extrapolation see property a priori/a posteriori see a priori/a
extreme realism 91, 131 see also posteriori
modal realism by acquaintance/direct 103, 105
independent of experience 62
fictionalism see modal fictionalism
formal language 6, 7 Lagadonian language 120
free will 7, 18 laws of nature 9
fundamentality 159 see also liar’s paradox 192
grounding linguistic ersatzism see ersatzism

9781472524263_txt_print.indd 222 01/12/2015 09:44

 INDEX OF TERMS 223

logical possibility 25, 29, 33 Quinian modal skepticism 50,

51–8, 70–1, 97, 114
master argument 34–5 radical 59, 71
material condition 27 modalism 13, 14, 74–90
mathematical entities 15 possible worlds and 80
maximality 130, 132, 134, 135, 136, monism
147, 195, 196 metaphysics 21
description 122 modality see modal monism/
external relation 98–100, 105, pluralism
109–12, 128
individual 95–7 naïve principle 188 see also
set 116, 119 Meinongianism
medieval modalitie 48 see also natural kind 8, 18, 99
scholastic philosophy on natural language 10, 11
modality nature 25 see also essence
Megarian school 31–5, 48 necessary connection 45
Meinongianism 106, 134, 176, necessary existents 133, 185–7, 194
187–90, 194, 196 necessitarianism 21, 41–2
mere possibility 59 necessity
mereology 94 see also composition biological 179
as identity concept 7
mereological complexity 98, 112 deontic 176
metalanguage 54 logical 176
metaphysical basis see philosophical metaphysical 179
method nomic 9, 15, 162, 176, 179
metaphysical monism see monism normative 176, 181
metaphysical possibility 9 physical 179
metaphysical necessity 15 varieties 15, 177–82, 193
metaphysical problem of possibility new actualism 15
see problem of possibility new essentialism see essentialism
metaphysics 8 nomic possibility 9
method see philosophical method nomic necessity 15
modal actualism nominalism 36–9
hardcore modal actualism 158 non-cognitivism 64
new modal actualism 156–72 non-contradiction, principle of 30
softcore modal actualism 158 norms 4 see also dodal normativism
modal agnosticism 14, 15, 149–54,
155 omnipotence 40–1
modal expressivism 12, 13, 64–70, 71 opinion see philosophical method
modal fictionalism 14, 118, 140–9, 155
modal logic 7, 71 paradoxes of Zeno 21–5
modal monism/pluralism 178, 180 philosophical analysis 6, 80, 197
modal normativism 67, 71 philosophical method 104–6
modal operator 10 philosophy of mind 8
modal realism 5, 14, 91–114 philosophy of science 9
modal reductionism 106, 109–12 placement problem 60
modal skepticism 7, 12, 13 plenitude of possibility 98, 129
Inwagen’s modal skepticism 50, possibility argument 58
57–9, 71 possibility concept 7

9781472524263_txt_print.indd 223 01/12/2015 09:44

 224 INDEX OF TERMS

possible worlds 4, 7, 13 realism vs nominalism see

cardinality 120, 134, 141 nominalism
contraction 129 recombination principle 111–12, 122
expansion 129 combinatorialism and see states
ideas and 43 of affairs
Leibniz and 42–4 humean recombination 98
possible-worlds semantics 4, 6, 7, 8, reductionism see modal
13, 42, 51, 87, 90, 100, 140–9, reductionism
164, 184, 186, 190, 191, 197 responsibility 9
powers 25, 28 see also capacities;
dispositions scholastic philosophy on modality
principle of humean recombination 36–9
see recombination sea battle example 30
probability 4 semantic interpretation 81
problem of possibility 3–5 semantic problem of possibility see
epistemic problem of possibility problem of possibility
3–5 sentiments 4, 64
metaphysical problem of possibility separability 45–6
3–5 situation 12
semantic problem of possibility skepticism 49–50 see also modal
3–5 skepticism
property 96 skeptical doubting 50
categorical/dispositional 165 see skepticism academic 50
also dispositions skepticism epistemic 50, 57
extrapolation 127 skepticism ontological 50
interpolation 127 state of affairs 123–30, 133–4
nuclear/extra-nuclear 189 see also constituents 124–6
naïve principle recombination 126–8 see also
universal 123, 129 recombination principle
propositional dictums 37–9 statistical or temporal frequency 27
propositions 8, 31–2, 131–3 substances 25, 36, 41, 43
de re propositions 37 see also de super-essentialism see essentialism
re modality surrogate see ersatzism
de dicto propositions 37 see also synchronic possibility 25, 29
de dicto modality
de se propositions 108 temporal modalities 9, 10
“that” clause 37–9
quantification 55 theodicy 42, 44
quantified modal logic (QML) 6, 51, thought experiment 40
54, 67, 75, 76, 80, 81, 82, 83, trans-world identity 138
86–90, 115, 116, 122, 130, 158, truth conditions 61
159, 161 truth maker 130
quiet moderate realism 131 see also
ersatzism, pictorial universal see property
Quinian modal skepticism see modal
skepticism varieties of necessity 15

real definition 83 Wittgenstein world 125–6, 128

9781472524263_txt_print.indd 224 01/12/2015 09:44