Ontology and Metaphysics
Ontology and Metaphysics
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
Thomas Hofweber
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
3
Great Clarendon Street, Oxford, OX DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Thomas Hofweber
The moral rights of the author have been asserted
First Edition published in
Impression:
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
Madison Avenue, New York, NY , United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number:
ISBN ––––
Printed in Great Britain by
Clays Ltd, St Ives plc
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
for Rebecca
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
Contents
Preface xi
Acknowledgments xv
How to Read This Book xvi
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
viii contents
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
contents ix
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
x contents
Bibliography
Index
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
Preface
This book hopes to make progress on a number of metaphysical problems that are
closely tied to ontological questions. The key to making progress in this area, I will
argue, is to have a better understanding of ontology itself: what questions it is supposed
to answer, how these questions can be answered, who should answer them, and for
what the answer matters. Our main focus will be ontological questions about natural
numbers, ordinary objects, properties, and propositions, and the corresponding,
larger metaphysical debates about the philosophy of arithmetic, the metaphysics of
material objects, the problem of universals, and the picture of reality as the totality of
facts. An overriding concern related to solving these problems will be what ambitions
metaphysics can have in answering questions about what the world is like. In the end
I will defend that metaphysics has more work to do than most philosophers think, but
less than most metaphysicians think.
In outline, this book proceeds as follows:
Chapter discusses four examples of metaphysical debates that are closely tied to
ontological questions. These four examples will be our main examples throughout:
natural numbers, ordinary objects, properties, and propositions. It then raises three
puzzles about how we should understand these ontological questions and how meta-
physical problems could possibly be so closely tied to them. Our overall goal will be
to solve the metaphysical problems by understanding the puzzles that ontology gives
rise to.
Chapter starts to solve the first puzzle by looking at the relationship between
ordinary non-metaphysical statements and apparently equivalent ones that are about
numbers, properties, propositions, truth, and other metaphysically loaded things.
These pairs of statements play a crucial role in trivial arguments for the existence of
numbers, properties, and propositions, and solving the first puzzle is closely tied to
understanding these arguments.
Chapter discusses quantification in natural language and how it is tied to onto-
logical questions. I propose, on the basis of considerations about ordinary communi-
cation, that quantifiers are polysemous, with one reading that is tied to ontology and
another reading that isn’t. This second reading is spelled out in more detail and the
idea is generalized to generalized quantifiers in an appendix to the chapter. At the end
of this chapter we will be in a position to solve the first puzzle.
Chapter outlines a strategy for answering ontological questions. It distinguishes
internalism from externalism about a domain of discourse, and spells out why it is
crucial to find out which one is correct. This chapter concludes the first part of this
book. After having seen what needs to be done to make progress, it is now time to try
to do that for our four cases.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
xii preface
Chapter starts with the first case: natural numbers. It investigates what we do when
we talk about natural numbers in ordinary conversations as well as in mathematics.
A proposal is made about why number words systematically play a double role in
natural language and what their semantic function is. Chapter attempts to solve
various problems in the philosophy of arithmetic in light of the account of talk about
natural numbers developed in chapter .
Chapter focuses on the metaphysics of ordinary objects, our second case. It
considers what we do when we talk about objects and defends that we have good
empirical reasons for an affirmative answer to the ontological question about ordinary
objects.
Chapter investigates talk about properties and propositions, our remaining two
cases, as well as some general issues in semantics. Chapter deals with the most
important objection to the account proposed in chapter . Chapter considers
whether there could be ineffable facts and how reality, understood as all that is the
case, relates to reality, understood as all there is. Chapter attempts to make progress
on the problem of universals and on metaphysical questions about properties more
generally. This concludes the second part of the book, which deals with ontological
questions and the metaphysical problems they are closely tied to. In this part a solution
to the second puzzle about ontology will be proposed.
In the third, and final, part we return to ontology and metaphysics more generally.
In chapter I try to find out whether ontological questions should properly be seen as
metaphysical ones, and how metaphysical questions relate to questions in other parts
of inquiry. Here we will see a solution to our third and final puzzle about ontology.
Chapter criticizes approaches to metaphysics that hold that metaphysics has a
unified subject matter that is, in one form or another, tied to what is fundamental or
some kind of metaphysical priority. A brief concluding chapter sums up and looks
ahead.
The position defended in this book, and the solutions advocated for the various
metaphysical problems, are based on a number of core ideas and theses, which will be
more fully articulated, explained, and defended in the chapters to come. They include:
. The polysemy of quantifiers. Quantifiers in natural language have more than
one reading, and we have a need for each of them in ordinary everyday
communication.
. The non-referentiality of certain singular terms. Expressions that syntactically are
like names can have very different semantic functions from names.
. Reference free discourse. Some domains of discourse systematically do not involve
any attempts at reference at all.
. Rationalist arithmetic. A rationalist philosophy of arithmetic based on an
empirical defense of natural number talk as a reference free discourse.
. A world of only objects. There are material objects, but no properties, no facts,
and no natural numbers.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
preface xiii
. The effability thesis. Every truth can be stated by us, in our present language.
. Conceptual idealism without ontological idealism. Reality, as the totality of things,
is independent of us, but reality, as the totality of facts, is not.
. A domain for metaphysics. Some questions of fact are properly addressed in
metaphysics, and some ontological questions are among them.
. Autonomy without freedom. Ontological questions in the domain of metaphysics
all have the same answer.
The view defended here draws upon some ideas put forth in various papers I wrote
over the last several years. The present book is an attempt to put these ideas together, to
develop them properly, to show how they are connected, to work out a coherent picture
of metaphysics and the role of ontology in it, and, in particular, to make progress on a
series of metaphysical problems. A few chapters are based on previously published
papers, but most are new. In particular, chapter is based on Hofweber (a),
chapter on Hofweber (a), chapter on parts of Hofweber (a), and parts
of chapter overlap with parts of Hofweber (a). In each case the material has
been updated and revised. The remaining nine chapters are new, although on a few
occasions a smaller section overlaps with a section from a previously published paper,
and sometimes ideas defended in the new chapters come up here or there in earlier
work. In particular, chapter is a much more detailed development of the criticism
of esoteric metaphysics outlined in Hofweber (a), and chapter does the same
for the view of quantification defended in Hofweber (b) and Hofweber (b).
The overall setup is inspired by Hofweber (b). Of course, no knowledge of any of
these papers is presupposed in the following.
All the readers of drafts of this book agreed on two things: first, it is too long for
anyone to be expected to read it all the way through, and, second, I really need to
say more about X (although there was a wide disagreement about what X should be).
I agree with this as well. I didn’t intend this to be a long book. To the contrary, I
planned on a short and snappy one, but it didn’t turn out that way. It became clear to
me that to do this properly I needed to start from scratch and motivate the problems
I hope to solve, to more properly develop the aspects of the overall view tied to
considerations about language, to include a discussion of all of the four cases that
are now included, to deal with the most important objections, and so on. Once the
manuscript got sufficiently long, I had to make it even longer to make the individual
chapters reasonably self-contained. This adds up to a lot of pages overall, and to many
large-scale philosophical debates engaged in a single book. It is natural to think that
I am trying to do too many things in one book, but for the overall position to come
into view these topics need to be addressed side by side. It is crucial, it seems to me, to
discuss all four of our cases of metaphysical problems tied to ontological ones, to set up
the general puzzles about ontology, to properly develop the account of quantifiers, the
larger picture of metaphysics, and so on. They all need to be in one book for the story
to come together. There is a shockingly long list of topics which I had to cut from being
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
xiv preface
discussed here, several of which were nicely written up, or so I thought, but deleted
to save space. What remains is what I feel needs to be there. I know it is too long for
anyone to be expected to read it all the way through, but I can’t make it shorter without
losing the big picture or having big gaps in the argument. In fact, I can assure you that
if you read it all the way through you will agree with the others who have done so that
(a) it is too long, but (b) I really need to say more about X. And as I said, I agree with
this as well.
This book does, and I believe needs to, discuss many large-scale philosophical
topics together: the nature of metaphysics, ontology, the semantics of quantifiers,
focus and syntax, number words in natural language, the philosophy of arithmetic,
perceptual beliefs, the problem of universals, idealism, ineffable facts, and several
others. Collectively the literature on these topics is beyond vast and it would be
impossible to do justice to all of it. This book doesn’t aim to do the impossible. It just
can’t be done to argue in each case why this alternative view is wrong, or why that
objection will go nowhere, or how the present position is different from another one.
All these are worthwhile things to do, but it being impossible to do them all should
not prohibit us from talking about all these topics together. In my discussion of the
literature I focused on the cases that struck me as especially relevant and important.
I submitted what I thought was pretty close to a final version of this book to OUP
in . Since then it has gone through several rounds of revisions in response to
substantial comments from six referees and many other philosophers. During those
revisions I changed the overall setup, dropped some chapters, added some new ones,
rewrote various parts, while the larger view remained stubbornly the same. I have tried
to keep up with the literature that appeared in the meantime and to incorporate it in
the text when possible, but often all I could do is mention a major new book or article
in a footnote rather than discuss it in the detail it deserves, and I suspect for many I did
not even do that. The emphasis of the cited literature is consequently often less on the
very latest, but more on what was most vivid when the basic things were put in place.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
Acknowledgments
Over the years I received help from many people with this project. Some helped me to
get it started and to get it on the right track, some to wrap it up, some saved me from
clear mistakes along the way, some had great objections that led to improvements.
Some helped me for a long time, some only in a short email exchange or a single
conversation. My sincere thanks to all of you! There surely must be others, but for
each of the following I can still pinpoint where or how they helped, although I
won’t try to spell it out case by case. My thanks to Bob Adams, Jody Azzouni, Kent
Bach, Brendan Balcerak Jackson, Dorit Bar-On, Jefferson Barlew, Karen Bennett,
Johan van Benthem, Jason Bowers, Otávio Bueno, Shamik Dasgupta, Catharine Diehl,
Matti Eklund, Anthony Everett, Sol Feferman, Katharina Felka, Hartry Field, Krasi
Filcheva, Kit Fine, Peter Godfrey-Smith, Charles Goodman, Tristan Haze, Randy
Hendrick, Eli Hirsch, Andreas Kemmerling, Boris Kment, Joshua Knobe, Toni Koch,
Dan Korman, Matt Kotzen, Robert Kraut, Phil Kremer, Wolfgang Künne, Marc Lange,
Pat Lewtas, Bill Lycan, Matt McGrath, Friederike Moltmann, Ram Neta, Jill North, Jeff
Pelletier, John Perry, Jesse Prinz, Michael Raven, Agustín Rayo, Mike Resnik, Tobias
Rosefeldt, Richard Samuels, Jonathan Schaffer, Kevin Scharp, Stephen Schiffer, Ben-
jamin Schnieder, Robert Schwartzkopff, Ted Sider, Allan Silverman, Keith Simmons,
Peter Simons, Rob Smithson, Rich Thomason, Amie Thomasson, Martin Thomson-
Jones, David Velleman, Lisa Vogt, Warren Whipple, Al Wilson, Julia Zakkou, Ed Zalta,
and surely several more.
I owe a special debt to six anonymous referees whose comments together were the
length of a decent size book and which led to many substantial changes. My thanks
to Peter Momtchiloff for finding these terrific referees, somehow persuading them to
take on this task, and his long-standing support of this project.
Some parts of this book are based on material published previously, as detailed in the
Preface. I would like to thank Duke University Press and Oxford University Press for
permission to reuse the material. Thanks to the Philosophers’ Imprint for not requiring
any permission to reuse one’s work. My thanks also to those who volunteered their
time and skill to give us LATEX and TexShop, which were used to write this book.
Finally, special thanks to Rebecca Walker for lots and lots of all kinds of things.
I dedicate this book to you, with gratitude for our life together.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
Obviously, the proper answer is: cover to cover. It is written with one main line of
argument in mind, starting with the motivation of a problem, and ending with a
proposed solution, going over four main cases in the middle, each of which has an
important role in the overall story. The next best thing would be to read the early
chapters – and then to read the chapters on just one of our four main cases, before
returning to the general discussion of metaphysics in chapter . Unless you have a
particular interest in one of the four cases, I think it might be best to focus on the
case of arithmetic and thus chapters and , skipping chapters –. Anyone who is
not tempted to understand metaphysics as being concerned with fundamental reality
or metaphysical grounding can also skip chapter , which criticizes these approaches.
So, a second best route to the main ideas of this book would be to read chapters – and
then chapter . If that is still too much, I would recommend reading a chapter on the
topic that interests you most. I put in an effort to make the discussion of the various
separate topics reasonably self-contained, and so I hope individual chapters can be
read by themselves with some benefit. Thus anyone who would like to focus simply
on quantification (chapter ), the philosophy of arithmetic (chapter ), the problem of
universals (chapter ), the ineffable (chapter ), the critique of esoteric metaphysics
(chapter ), or any other chapter, will hopefully find that to be a possibility. A few
chapters have one or more appendixes that deal with further issues from that chapter.
They are important, but they can be skipped without loss of the overall argument by
anyone satisfied enough with the treatment of the issue in that chapter. But before
spending too much time on deciding on which parts to read and which parts to skip,
let’s just start from the beginning and see where it goes from there.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
Ontology and Metaphysics
. Introduction
The grandest and most ambitious part of philosophy is metaphysics: the project
of finding out, within philosophy, what reality is like in general. A central part of
metaphysics is ontology: the project of finding out what kinds of things make up
reality, what exists, or what there is. Metaphysics could thus roughly be divided into
two parts. The first is ontology, which is supposed to tell us what there is in general, or
what kinds of things make up reality. The second is the rest of metaphysics, which
is supposed to tell us, among others, what these things are like in various general
ways. Many of the great philosophical problems are problems in metaphysics. And
many of these are closely tied to problems in ontology. We will see some examples
shortly.
There also is a long tradition in philosophy to find metaphysics and ontology
suspect. Prominent worries are that metaphysical statements are meaningless, or that
metaphysical investigations won’t lead to knowledge. But such worries are usually
based on a very strict requirement for something to be meaningful or to be knowledge,
a requirement so strict that most other things wouldn’t be meaningful or knowledge
either. I don’t find much to worry about in these objections, but there is another
worry which I think is terribly serious. It casts doubt on metaphysics in general, or
at least suggests we don’t quite understand what we are supposed to do when we
engage in metaphysics. This worry isn’t that the question isn’t meaningful or that the
answer can’t be known, but rather the opposite: the questions are perfectly meaningful,
but the answers are already known. There is no work for metaphysics to do, the
worry goes, since whatever work there is supposed to be done, it has already been
done elsewhere. The allegedly deep questions that metaphysics, in particular ontology,
hoped to answer have long been answered and their depth was an illusion. The depth
was an illusion not because the questions are somehow defective, or in some way
relative, but because they are perfectly non-defective and objective questions that
have long been answered rather trivially in other parts of inquiry. A prime example
of this situation, to be discussed in detail shortly, is the ontology of mathematical
objects. The metaphysician seems to want to know whether there are any mathematical
objects: things like numbers. But the mathematician has proven long ago that there are
infinitely many prime numbers, and thus that there are numbers.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
This worry, once properly spelled out, suggests that the project of metaphysics and
ontology isn’t as well defined as one should hope. In light of this worry one might be
tempted to give up on metaphysics, but this would be premature, since there are two
promising ways to defend it. One way to it is to hold that metaphysics has a distinct
and unified subject matter that is sufficiently separate from other parts of inquiry.
Metaphysics is not supposed to find out what reality is like in general after all, but
instead what reality is ultimately or fundamentally like. And these notions are to be
taken in a metaphysical sense, which makes sure that those questions aren’t trivially
answered in other parts of inquiry. A second way to defend metaphysics is to hold
that once we pay close attention to what we try to do when we do ontology and
metaphysics, and how we articulate the questions we hope to ask, then we can see
that these questions are neither trivial nor already answered. This second approach
doesn’t require a distinct subject matter for metaphysics, but instead a careful study
of the language we employ and the goals we have when we ask the relevant questions.
Both approaches defend metaphysics, but they do so in different ways. Both hold that
metaphysics can have real ambitions to contribute to inquiry, but their ambitions will
be different. Whichever defense one adopts will lead to rather different versions of
what the discipline should aim to do, and how its problems are to be solved. We will
discuss both approaches in detail, but I will defend the second one in this book.
Thinking about metaphysics as a discipline is often an anti-metaphysical activity.
Many philosophers who wondered about metaphysics as a whole did so in order to
reject metaphysics, with the logical positivists being the prime example. But this does
not have to be so. There are real puzzles about metaphysics and ontology, and their
discussion needn’t be tied to attempts to reject metaphysics. Those fond of metaphysics
often react to questions about metaphysics in general negatively, taking them to be
an attempt to demand some outside justification for a practice that is doing well by
its own standards. Their motto is: don’t ask what metaphysics is supposed to do, or
how it is possible to do that, just do metaphysics. But both of those attitudes are
mistaken. Thinking about problems for ontology and metaphysics in general is not
anti-metaphysical, but instead the key for answering some of the most important
metaphysical and ontological questions, or so I hope to argue in this book. The goal
in the following will be to show that we can make progress on a number of significant
metaphysical problems that are tied to ontology by first thinking about metaphysics
and ontology in general. I will argue that a crucial obstacle to finding a solution to these
problems has been a mistaken conception about what ontology is supposed to do
and how it is supposed to be done. Once we correct this error we will be able to
see the answer to several ontological questions, and consequently to a series of larger
metaphysical questions.
Throughout this book we will focus on four ontological problems which are closely
tied to four large-scale metaphysical debates. These problems are our focus because of
their significance within metaphysics and because they bring out a variety of different
ways a metaphysical problem can turn out. As we will see, two of them are similar to
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
each other, but the other two are different from all the others. What we will conclude
about these four ontological problems in particular can’t be expected to carry over to
any other ontological problems, but what we will conclude about ontology applies to
ontology in general. But before we can get there we have to briefly reintroduce the
ontological problems we will focus on, as well as the metaphysical questions they are
tied to.
.. Numbers
A good example of the connection of large-scale philosophical questions to issues in
ontology is the philosophy of mathematics. Mathematics is philosophically puzzling
in a number of ways. First of all, it is a discipline that is highly objective. The results it
achieves should be accepted by everyone, no matter what one’s cultural background,
preference, or opinions. Mathematics is not alone in this regard, of course. Other
disciplines are equally objective. The natural sciences, which try to find out various
things about the natural world, often have that feature as well. And this suggests that
mathematics might be objective for a similar reason that the natural sciences are
objective. The latter aim to correctly describe a certain aspect of the natural world,
which in turn exists independently of us and our opinions. In this sense particle
physics aims to correctly describe subatomic particles, biology aims to describe living
things, and so on. These are all objective disciplines since the objects they aim to
describe, as well as their features, exist independently of us and our opinions. But if
mathematical objectivity is understood along these lines, then what part of reality does
it aim to describe? It is tempting to think that it aims to describe its own, distinct part
of reality: mathematical objects. These objects would likely be neither in space nor
time, and quite different from ordinary objects we normally interact with. But if they
exist independently of us and mathematics aims to describe them, then no wonder
mathematics is so objective.
On the other hand, the method with which mathematics achieves its results seems
to be in prima facie tension with it having mathematical objects as its subject matter.
Mathematics achieves results paradigmatically with proofs, and these proofs can
in general be carried out with one’s eyes and ears closed, by just thinking. But if
mathematics is about some objects that exist independently of us, how can we find
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
out about them by just thinking? There might thus be a better way to think of the
source of mathematical objectivity, one that can live with there being no mathematical
objects, and the following is one way this might go. When we carry out the proofs we
all rely on certain assumptions, certain principles, or axioms, at least implicitly, that
characterize the subject matter about which we hope to carry out proofs. But wouldn’t
those assumptions alone be enough to account for the objectivity of mathematics?
Wouldn’t it be enough if we all shared the assumptions that are the starting points
of our proofs? If the objectivity can come from our shared assumptions, then the
mathematical objects might simply fall out of the picture. Mathematics thus might not
aim to describe a certain part of reality after all, but do something else. And whatever
else that is more precisely, it might make clear that thinking alone is enough to achieve
mathematical results, that mathematics is fully objective, while nonetheless there are
no mathematical objects that mathematics aims to describe.
These are two large-scale views in the philosophy of mathematics that are closely
tied to the ontological question of whether or not there are any mathematical objects
in the first place. If reality contains some special part of mathematical objects, then
this is what mathematics will have to capture correctly in order for its results to be
true. But if there are no such objects, then this is not how mathematics and the source
of its objectivity are to be understood. A crucial dividing line in the philosophy of
mathematics is thus what one says in answer to the question
() Are there mathematical objects?
If one says “yes” one will be in one camp, if one says “no” one will be in a different one.
In the following, for reasons that will become clear later, we will in particular
focus on the philosophy of arithmetic, the mathematical discipline that deals with the
natural numbers. Here the general questions are just the same, but the mathematical
objects, if there are any, that arithmetic would be about are special objects: natural
numbers. A crucial question for the philosophy of arithmetic is thus the question
() Are there natural numbers?
However one answers this question will put one in one of two camps about how to
understand the philosophy of arithmetic. Progress in the philosophy of arithmetic is
thus tied to finding the answer to this ontological question.
.. Ordinary Objects
The objects we interact with most in everyday life are artifacts like tables and chairs,
biological organisms like human beings and trees, and inanimate natural objects like
rocks. All these are midsize objects, which can in principle be divided up into smaller
pieces from which they are made up or composed. Even though we think of the world
as containing such objects, there are a number of worries about whether this is indeed
correct. One of them is tied to how the midsize objects relate to the things from which
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
they are built up. All these objects are made up from smaller parts like atoms. But do we
really have good reason to think that there are these objects besides the atoms, or is all
there is just the atoms arranged in certain ways? The main alternative to the standard
view that there are ordinary objects is some version of atomism, that is, a view that
holds that there are just the simplest things, but nothing that is built up from them, in
particular no tables. There are some good arguments that suggest a form of atomism is
true: doesn’t science support that all there is is the subatomic, physical world? Don’t the
atoms do all the work when it comes to causing things, including our experiences, and
so anything in addition to them, like tables, are just extra baggage? Isn’t it somewhat
arbitrary and mainly due to us which collections of atoms we call “an object” and thus
put together into a unit? Doesn’t thinking about the world in terms of ordinary objects
reflect more on our thinking than on how the world actually is? Whether the atomist
or the ordinary conception of the material world is correct seems to be closely tied to
an ontological question, the question
If the answer is “yes”, then one large-scale picture of the material world is correct; if
the answer is “no”, then a very different one is correct.
.. Properties
The world, as we commonly think of it, contains at least ordinary objects. Whether
this conception of the world is accurate in the end is up for dispute, as we just saw
above, but even if it is correct there is a further question about whether or not that
is all there is. Are there just particular things, or is there more? One of the most
compelling arguments for there being more is simply the following argument. The
particular things are a certain way. Some are red, some are green. Those that are
red have something in common: being red. So, there is something all red things
have in common. But what is that thing? Not another particular red thing, which
has a particular location, but something different from and separate from all the
red things. It is a property or universal. It is something that all the red things
share, that they partake in, or somehow relate to. Thus reality contains more than
just particulars, it also contains universals which those particulars share, or so the
argument goes.
But if there really are properties or universals beside individuals, then where are
they and how do they relate to the individuals that have them? To say how this
relationship is supposed to go has been notoriously difficult for those who believe
in such universals. If the universal is where the thing is that has it, then it seems
that universals are at more than one place at a time. If the universal is somewhere
else, maybe outside of space and time, then it seems to leave the object by itself as a
completely featureless thing or bare particular. And it leaves the question open how
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
these bare particulars relate to the universals which are somewhere totally different.
Maybe then the world consists just of individuals, and not also properties that they
have in common. But does that mean that individuals have nothing in common? Do
red roses not all have one thing in common: being red, that is, the property or universal
of being red?
This group of questions is one of the oldest and most widely discussed problems of
all of metaphysics: the problem of universals. There are subtly different versions of what
the problem is more precisely, but in a nutshell it is the problem of saying (a) whether
there are besides individuals also other things which are the properties that they have,
and (b) if so, what these properties are, where they are, and how they relate to the
individuals that have them.
The problem of universals is related to another one, which seems even more
important. It is the problem of whether there is an objectively best description of the
world in the following sense. We describe the world with sentences that are mostly of
a subject–predicate form. In the simplest case the subject term picks out an individual
that we are making a claim about, and the predicate specifies what we claim of the
individual. So, “Fido is a dog” says something of Fido, namely that he is a dog.
Predicates group things together. There are many other things besides Fido of which it
can be truly said as well that they are a dog. But among all the different ways in which
things can be grouped together by the predicates we use in our language, are there
some that are objectively better than others, ones that the world itself suggests, not
simply that we prefer? This is, of course, a little metaphorical, but one could see how
it might be made sense of by talking about properties. If there are properties, if the
world contains them beside the individuals that have them, then it might seem that
the world recommends our grouping things together with predicates the same way it
groups them together with properties. If there is a property of being a dog, one that
groups all and only the dogs together, then it might be a good idea to have a predicate
in our language that expresses just that property, and that groups things together in the
same way. In particular, maybe only some predicates express properties, which would
make these predicates special. With predicates we can group things together any way
we want, but properties are how the world itself groups things together. But if there
are no properties, if the world contains just individuals, then it would seem that it is
up to us, and us alone, to group things together. The world remains silent on which
things go together, we can take our pick.
Whether the world we live in is a world of only individuals and whether there is a
distinguished description of it are questions that are closely related to the question of
whether or not there are any properties besides the individuals. One crucial dividing
line in understanding the world of individuals is thus what the answer is to this
question:
() Are there properties?
This is a question in ontology and one of central importance for metaphysics.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
.. Propositions
It is natural to hold that thoughts involve at least two things. One is the person, who
is thinking the thought. The other is the content of the thought, that is, what one is
thinking. When I think that Obama is the president then I am the thinker, and the
content of the thought is that Obama is the president. These contents are generally
also called propositions. Thus when I think that Obama is the president I have a
thought that has the proposition that Obama is the president as its content. Contents
or propositions give rise to two related philosophical debates.
First there is the question of whether thinking indeed involves relating to some thing
which is the content of one’s thought, and if so, how we mange to relate to such things
as propositions. Propositions presumably are not ordinary objects, so the question
remains how I manage to relate to this one when I have one thought and that other one
when I have a different thought. More importantly, one might think that this picture of
the contents of thoughts as relating to propositions is mistaken. An alternative would
be to hold that our thoughts have contents, but we do not relate to things which are
contents. So, when I think that Obama is the president I relate only to Obama, and
think of him that he is the president, not to some other thing which is a content. On
such a picture there would be no such things as contents to which we relate, even
though our thoughts have contents in the sense that they are contentful.
A second issue connected to propositions is not directly about the contents of
thoughts, but “the propositional” more broadly. The proposition that Obama is the
president is closely tied to the truth, or the fact, that Obama is the president. If there
are such things as propositions then it is not unreasonable to think that there are such
things as truths or facts. The world then would not simply contain objects and possibly
also properties, but in addition also facts or truths. And if the world contains such
things then this might suggest a certain form of realism: what can in principle be truly
said about the world is already contained in the world itself, as one of its building
blocks. To say something truly is just to say something which is a truth, and what
truths there are is something that is a part of what there is in general. The truths are
all there already, so to speak, and we can only hope to capture as many of them as
possible. But if the world contains only material objects, say, then there are no such
things as propositions, and no such things as facts or truths. We will still be able to
think truly about the world, and say true things about it, but the world itself does not
contain any truths or facts. And this suggests a rather different picture of what we do
when we aim to describe the world. This contrast might be made clear when we think
about whether it could in principle be that there are some truths that creatures like us
could never say or think. If facts or truths exist as part of the world, then they are just
there, waiting to be thought. It might well be that we can get all of them in the end, but
since they are just there, as an independent part of the world, it would be reasonable
to expect that we can not get them all. Some might well be completely beyond us, that
is beyond what creatures like us can think or say. On the other hand, if the world does
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
not contain any such facts or truths or propositions, then this might be very different.
What can in principle be thought is not directly tied to what propositions are part of
the world, and so it might be tied to us in a way that we can think everything there is
to think.
The questions how we manage to have thoughts with contents and whether we can
think all there is to think are tied to an ontological question:
() Are there propositions?
Whatever one says in answer to this question will lead to different views of the contents
of thought and its limits.
All these large-scale metaphysical issues discussed over the last couple of pages thus
seem to be closely tied to ontological questions. To solve these problems seems to
require an answer to the ontological questions. We will need to see in the following
whether this is correct. That is, we will need to see whether the ontological questions
indeed do have a central role in these debates, what this role might be, and, in
particular, how we should understand the ontological questions more precisely. As
we will see shortly, to give ontological questions a central role in metaphysics is
problematic, as are the ontological questions in the first place. To understand the
metaphysical problems outlined better we will need to understand ontology better.
And once we do that we will be in a position to answer these metaphysical questions,
or so I hope to show. But first we need to look at ontology more closely.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
be expressed, something we will discuss shortly. But a natural first try is simply this:
the primary ontological question about Fs is the question of whether there are any.
(POQ) The primary ontological question about Fs: are there Fs?
If the answer is “no” then this would seem to settle the issue and no further questions
about Fs would follow. But if the answer is “yes” then it seems natural to follow up with
a request for more information about what these things are in general like.
(SOQ) The secondary ontological question about Fs: what are Fs in general like?
Some philosophical problems in ontology are primary ontological questions, some
are secondary ontological questions. Sometimes the two questions get mixed together.
The problem of universals is often stated as the problem of saying whether there are
any universals, and if so how they relate to the particulars that have them. Much
of the work that is actually carried out is usually in the part that deals with the
secondary ontological question, in part to justify the answer one gives about the
primary ontological question. So, if one says “yes” to the primary question, then one
better have some answer to the secondary ontological question. And even though it is
easy to say “yes” to the primary ontological question, it turns out to be hard to answer
the secondary ontological question in a satisfactory way. Sometimes the secondary
ontological question comes first in the order of inquiry. We might first determine
what ordinary objects would have to be like, if there were any, and then use this
characterization to argue that there aren’t any, since nothing is or can be like that.
Secondary ontological questions also overlap metaphysics more broadly understood,
in the way that goes beyond ontology. Here there are not clear dividing lines, and it
is simply a matter of from which angle one approaches a set of issues. Still, we will
in the following at first focus mostly on the primary ontological questions, and only
return to the secondary ones a little later. It will turn out that a proper understanding
of the primary ontological question is what is central for the answer to both and for
metaphysics more broadly.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
when we try to do this we think that our understanding of mathematics is tied to what
we should think the answer to certain questions is, including the question of whether
or not mathematics aims to describe an independently existing domain of things—the
mathematical objects—which are the source of mathematical objectivity, the subject
matter of mathematics, and so on. And thus a crucial fork in the road is what one
says is the answer to the question of whether or not there are any mathematical
objects. And that question is intended to be part of the philosophical project of
trying to understand mathematics, to see what it looks like from a philosophical
point of view.
But this project seems to be completely incoherent. One can’t coherently accept
the results of mathematics, but still wonder whether there are mathematical objects.
Mathematical objects are paradigmatically things like numbers, functions, or sets. But
the results of mathematics themselves imply that there are mathematical objects such
as numbers, functions, or sets. Take Euclid’s Theorem, which was proven about ,
years ago and has been beyond question ever since. It states that
() There are infinitely many prime numbers.
This immediately implies that there are infinitely many numbers, and thus that there
are numbers. But how then can we ask whether there are mathematical objects, things
like numbers, without thereby also questioning the results of mathematics? It just
seems incoherent to have this philosophical stance, and to engage in the project of
trying to understand mathematics from the outside by asking such questions.
But on the other hand it seems perfectly coherent, and philosophers have been
doing it for quite a while with a strong sense of a meaningful project. They want to
know whether mathematics aims to describe a special part of reality: the mathematical
objects. And this is tied to the question of whether or not there are such objects in the
first place, whether or not reality contains besides the ordinary stuff in space and time
also other things that mathematics is about, things like numbers. What should we
think about this situation?
The situation is actually even worse for the philosophical project than I just made it
out to be. If we wonder whether or not there are any natural numbers, we don’t have
to rely on mathematics to conclude that there are such numbers. We can conclude this
already from the fact that I have two hands, and thus the number of my hands is two,
and thus there is at least one number: the number of my hands, i.e. two. Can I really
answer the ontological question about mathematical objects by drawing a couple of
trivial inferences from the fact that I have two hands? Even worse still, can’t I simply
conclude that there are numbers by giving examples of them: the number two is one,
the number three is another?
Ontology was intended to be part of a metaphysical project that asks a further
question, one that goes beyond the questions that are addressed, and often answered,
within a discipline or domain of discourse that the metaphysician tries to theorize
about. Metaphysics so understood was ambitious in that it had further facts to uncover
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
than those that were already investigated in the discipline it was trying to understand.
And these further facts were further in the sense that those from within the discipline
did not themselves settle the corresponding questions. There was really more work to
do, and ambitious metaphysics set out to do it. But how the ontological questions can
be seen as ambitious metaphysics in this sense is just what seems to be incoherent.
The question we philosophers want to answer is apparently already answered within
mathematics itself. Whether the philosophical project of ontology is coherent, what
question it is supposed to answer, and whether these questions can play an important
role in the philosophy of mathematics and metaphysics in general is what we hope to
make progress on in this book.
There are really several different aspects of this worry. They concern both the
philosophical project of asking ontological questions, as well as giving the answer to
these questions a significant role in one’s philosophical account of a certain discipline
or domain of discourse. There are issues about whether we expressed the question we
wanted to ask properly, how it could be that this question is answered the way it seems
to be answered, what the answer matters for, and why any of this should be seen as
answering a philosophical question. It is good to bring out the different aspects of this
issue more clearly. I will in the following isolate three puzzles that, as I will argue, are
especially important. These puzzles will be with us for most of this book, and to solve
them will require quite a bit of work in each case. The three puzzles are in essence
three different aspects of the worry outlined above. Not all of them apply equally to
all ontological questions, but the fact that each one of them applies to some important
ontological questions should give rise to the worry that we don’t understand ontology
until we understand each of them better.
in particular the ones considering these three cases. This question has two apparently
equally good, but contrary answers.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
such things. But figuring that out is at least as hard as figuring out what the best theory
of the world as a whole is. And most of that won’t be work for philosophy to do. So,
answering ontological questions is very hard.
answer : trivial
The questions of whether or not there are numbers, properties, or propositions can
all be answered trivially in the affirmative. There are trivial arguments that establish
that there are numbers, properties, and propositions, and these arguments use only
completely uncontroversial premises, and just a couple of easy steps of inference. Here
are some examples:
Of course, each of these arguments can be disputed. Each of them involves several
steps where a philosopher can, and has, come along and deny that it is a valid
inference. But let’s be honest: are there really any easier and more compelling argu-
ments anywhere in philosophy? Is there really anything we can establish more easily?
The questions of whether there are numbers, properties, or propositions are trivial,
the answer is obviously: yes. And that this is so can be seen, among other ways, by the
above trivial arguments. Nothing could be easier than to see that there are numbers,
properties, and propositions. These trivial arguments don’t carry over to all ontological
questions. For example, there seems to be no analogous argument for there being
ordinary objects, and we should try to understand where this difference comes from.
But that there are such trivial arguments for answers to some ontological questions is
enough to give rise to a puzzle about ontology.
1 Jupiter in fact has more than four moons. Nonetheless, I will use this example, as do many others, since
it is a classic example of Frege’s, from his Frege ().
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
But maybe that isn’t right. Maybe what we are trying to find out when we do ontology
and metaphysics is not whether there are such things, but something else? Maybe we
were misled into expressing the question we want to ask in a certain colloquial way, and
that misarticulated question was trivially answered. But how else should we express
the question we want to ask? Most naturally, we should look back at how ontology was
motivated. Here we used expressions like “make up reality,” or “exists,” and thus maybe
we should instead try to ask the question as follows:
2 See Chalmers () or Kraut () for such broadly anti-realist positions.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
or
() Do numbers exist?
or maybe simply
() Are there really numbers?
But it is not so clear how much progress that would be. The first question seems almost
a bit strange (what else would they be part of if there are any in the first place?), and
the second and third are not clearly different from what we had before. How could it
be that there are numbers, but they don’t exist, or that there are numbers, but there
aren’t really numbers? It is not clear that by putting things this way we get any more
substantial questions. Furthermore, couldn’t we have stated Euclid’s Theorem above
just as well as the result that infinitely many prime numbers exist? Such attempts at
a reformulation of the question might be on to something, but the above proposals
so far won’t be the solution. What we can conclude, however, is this: philosophers try
to ask certain questions that naturally get grouped together in the discipline we call
“ontology.” We can call these questions ontological questions. Ontological questions
generally are directed at certain cases, like numbers, objects, properties, or the like. We
can call them the ontological question about numbers, the ontological question about
objects, and so on. What we have to find out is thus this: what are the ontological
questions (about numbers/properties, etc.)? That is to say, how should we properly
express these questions that we want to ask? To make progress on this we should in
particular see whether or not the trivial arguments are indeed valid. If they are not
valid, then maybe the question we wanted to ask is properly stated as (). But if they
are valid, then it would seem that either ontology is trivial (for these cases), since ()
is trivially answered, or the question has to be stated in some other non-trivial way, or
some form of anti-realism is true about ontology. To solve the first puzzle requires in
part to find out what question we are trying to answer when we do ontology.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
the most important part of inquiry. Whatever one in the end thinks of the value of
a moral philosophy, or political philosophy, metaphysics won’t have that importance,
and ontology certainly will be even less important. Even those among us who work in
metaphysics see some justification for the lack of multi-million dollar research funding
for our field, even though such sums are frequently given to such projects as finding
out how viruses attack cells, what subatomic particles there are, the Hubble Space
Telescope, or even finding out how to factor large numbers. Philosophical questions
might be questions that force themselves on to all human beings, but pursuing them
is a luxury that we are fortunate enough to be able to afford. But like all luxuries, they
are not very important.
Fodor has a point here. If it turns out that we don’t have beliefs and desires, and thus
the proper explanation of what we do does not involve them, then this would shake our
conception of ourselves in a profound way. It really would be an intellectual disaster.
But what if it turns out, in ontology, that there are no propositions? Then there are
no such things as contents. In particular, there is no such content as the content that
snow is white. But if there is no such content, how can anyone have a thought with
the content that snow is white? You can’t, literally, have a thought for which its content
is missing. But all beliefs and desires, the building blocks of belief–desire psychology,
have to have contents. If there are no propositions, then there are no contents, and
thus there are no beliefs and desires. And so belief–desire psychology cannot be true.
If ontology goes one way, the greatest intellectual catastrophe of our species is right
around the corner.
Motivating the threat of disaster this way is, of course, not philosophically innocent.
For example, there are attempts to separate the ascription of beliefs and desires with
taking them to be relations to contents or propositions.3 But the threat of a disaster
is quite real since such approaches are generally considered to be a minority view,
3 Some examples of such theories are the syntactic relational theory of Davidson () and the multiple
relation theory of Russell () and Moltmann (b).
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
one which might well turn out to be incorrect. If belief relates us to propositions, as
it seems to, then the negative answer to an ontological question leads to the greatest
intellectual catastrophe in human history, as Fodor put it. Ontological questions can
thus be extremely important.
Things are even worse when it comes to ordinary objects. If it turns out that there are
no ordinary objects, then not only are there no tables and chairs, there are none of us
either. If all there is is just atoms in the void then we don’t exist, since we are not atoms.
Even though the non-existence of propositions might lead to the greatest intellectual
catastrophe in human history, the non-existence of ordinary objects would be worse,
since it guarantees that there are no humans and thus there is little human history.
And similarly, although less dramatically, for numbers. If it turns out in ontology
that there are no numbers after all, then there are no prime numbers, and no even
numbers, and no odd numbers, and so on. And if there are no prime numbers, then
it is false that there are infinitely many prime numbers, something Euclid thought he
established with his famous proof. And similarly almost all of mathematics is false.
And if mathematics is false, then so is a lot of physics and economics and biology,
since they all rely on mathematics. Thus if the ontological question about numbers
goes one way, then everything else goes down the drain. The ontological question is
thus of unparalleled significance. Everything depends on it.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
The second puzzle is also closely related to the question of the source of the
objectivity in a particular domain of discourse. If arithmetic, for example, is about
an ontology of mathematical objects, the numbers, then they could be the source
of mathematical objectivity. They might exist independently of us, and adjudicate
impartially all mathematical disputes. But if our arithmetical talk in everyday life
as well as mathematics does not require an ontology of numbers, and if its truth is
independent of their being such an ontology, then it won’t be the source of arithmetical
objectivity. To solve the second puzzle requires us to also make headway on this issue.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
science, not philosophy. And similarly for there being ordinary objects, like bars of
metal. The question of whether or not there are numbers, objects, or properties is not
a philosophical one at all. It is one that has long been settled in the sciences.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
It should seem promising that if we could solve these three puzzles, and understand
how they arise, then we would be in a good position to find out what role ontology
should play in metaphysics, and whether metaphysics can be ambitious in the way
we hoped: that it asks further questions, not immediately answered in other parts of
inquiry, but still significant questions of fact. A promising strategy to achieve this is
thus to look at the three puzzles, to see how they arise and to try to solve them. This
is the strategy I will pursue in this book. I hope to make the case that it is the right
strategy by showing that it indeed bears fruit: a solution to these puzzles illuminates
not just ontology, but also the extent of the proper ambitions of metaphysics and the
place of ontology in it.
The three puzzles increase in difficulty in the order given above, and a solution to
the later ones will depend on a solution to the earlier ones. We will need to solve the
first puzzle first, before we can attempt the second one, and we will need to solve that
one before we can move on to the third one. We need to know first what the ontological
questions are before we can see how important they are, and what depends on their
answer. And only after we know that can we hope to see what the philosophical project
can be that is connected to them.
I will thus start with the first puzzle. To understand how it arises we should look at
the trivial inferences that apparently allow us to conclude from ordinary statements
that there are numbers, properties, and propositions. These trivial inferences can be
broken down into two parts. The first is the part where we conclude, for example, that
the number of moons is four from the premise that there are four moons. The second
is the part where we draw the conclusion that there is a number from that the number
of moons is four. Both of these inferences deserve a rather detailed discussion. This
will be our point of departure for the next two chapters: chapter will discuss the
first inference, chapter will discuss the second one. At the end of chapter we will
be able to solve the first puzzle, and we will have made substantial progress towards
understanding how to articulate an ontological question properly. It will then be time
to move on towards solving the second puzzle and to understand how ontological
questions relate to the objectivity of a domain of discourse. The middle part of this
book will deal in some detail with what we do when we talk about numbers, objects,
properties, and propositions. We will see what the solution to the second puzzle
is for our four cases by the end of chapter , and it will be different for different
cases. It will then be time to think about the status of ontology as a philosophical
discipline, and about how metaphysics relates to other parts of inquiry. We will see
the solution to the third puzzle in chapter . By then I hope to have made the case
that metaphysics can indeed have the ambition to ask further questions of fact, that
these questions sometimes are ontological ones, that they are, just as we originally
thought, just questions like “Are there numbers?,” what the answer to that question is,
and why it all matters. But in order to get there we need to first try to solve the first
puzzle.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
Innocent Statements and Their
Metaphysically Loaded
Counterparts
which are taken to be deep, substantial, and important philosophical questions. But,
on the other hand, it seems that there are trivial arguments that immediately imply an
answer to that question. In particular, it seems that this question can be answered so
trivially that even if one wasn’t surprised by the answer, one should be surprised by
just how trivial it was to get it. There are arguments that answer question () in the
affirmative that rely only on completely uncontroversial premises, and they establish
this conclusion in two apparently completely trivial steps. These arguments are the
main topic of this chapter. There are a variety of lessons that could be drawn from
these arguments. One is that the question () is indeed trivial, but it is not the one
that we should ask in ontology.2 Or one might hold that the arguments are not valid.3
Or one might hold that ontology is trivial after all.4 Or one might hold that the whole
project of ontology is confused in some way.5 I will draw a different conclusion below.
To see what one should conclude from these arguments, we will have to look at them
in more detail. In doing this we will have to start investigating what we do when
we talk about numbers, objects, properties, and propositions. We will at first, in this
chapter, only look at a very limited range of cases of such talk, a range that will expand
significantly in later chapters. The cases we will look at now are especially significant
1 This chapter is based on Hofweber (a). The overlapping material in it has, however, been revised,
updated, and reorganized.
2 See, for example, Fine (), Dorr (), Fine (), Schaffer (b), and many others.
3 See Field (a) or Yablo ().
4 See, for example, Schiffer () for properties and propositions, Frege () and Wright () for
numbers, and Thomasson () for ordinary objects, and Thomasson () for ontology more broadly.
5 See, paradigmatically, Carnap ().
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
innocent statements
for the question of how to solve the first puzzle and how to understand the trivial
arguments. We will look at the general case in the chapters to follow, which will be
central for solving the second puzzle about how important ontology is.
The simplest and most striking arguments that there are numbers, properties,
and propositions, all have the same form. They start out with only one premise: a
completely trivial and everyday statement that seems to have nothing to do with
metaphysics. I will call such statements innocent statements, since they are apparently
completely free of any metaphysical baggage. Examples include:
and basically every other ordinary, everyday statement. The argument then moves to
one or more of the what I’ll call metaphysically loaded counterparts of the innocent
statements. These are statements that are apparently equivalent to them, and obviously
so, but that introduce things that don’t seem so metaphysically innocent any more:
numbers, properties, propositions, and truth:
Once these are established it seems to require only a further trivial inference to bring
in quantification. Once we know that Fido has the property of being a dog then we
know that there is a property that he has, namely being a dog. So, after the inference
from an innocent statement to one of its metaphysically loaded counterparts a second
step takes us from the metaphysically loaded counterparts to a quantified statement:
() There is a number, which is the number of moons of Jupiter, namely four.
And once we have established that, it seems completely trivial to conclude that there
are numbers, properties, and propositions.
If the ontological question about numbers just is the question of whether there
are numbers then it would seem that this ontological question is indeed trivially
answered. Before we draw any conclusions about what the ontological questions really
are, or whether ontology is trivial in these three cases, let’s have a closer look at
the trivial arguments. These arguments are of interest quite independently of the
6 Examples that are used frequently will receive mnemonic labels, to make their identification easier
below. Other examples will simply be numbered.
i i
i i
i i
OUP CORRECTED PROOF – FINAL, //, SPi
i i
innocent statements
larger issues about the status of metaphysics, since they do give rise to a number
of puzzles that are independent of metaphysics. These are puzzles about language,
and understanding them, I will argue, is the key to understanding the puzzles about
metaphysics. Stephen Schiffer has called the inferences from innocent statements to
their metaphysically loaded counterparts something-from-nothing transformations in
Schiffer (). Whether they involve a transformation of some kind or merely an
inference is up for debate, but somehow we seem to be able to infer, from nothing, that
there are such things as numbers, properties, or propositions. The question is whether,
and how, we can get something from nothing.
We need to carefully look at the trivial arguments to make progress on this issue,
and we will do so one step at a time. First we will investigate the inference from an
innocent statement to one of its metaphysically loaded counterparts in this chapter.
Then we will look at the second step of the trivial arguments in chapter : the inference
from the metaphysically loaded counterpart to a quantified statement. After that we
will be in a position to solve the first puzzle.
A The trivial arguments are not valid. They appear valid, however, for one of
several reasons.
Among those reasons that make the arguments appear to be valid, even though they
are not, could be a connection to fictionalism. If the metaphysically loaded statement is
not literally true, but only true given that there are numbers/ properties/propositions
at all, then this might well explain why the argument seems to be valid. We might slide
between what is literally true, the innocent statement, and what is only true given the
fiction that there are numbers. A version of this was defended by Hartry Field in Field
(a), and by Stephen Yablo in Yablo ().7
On the other hand, one might think that the trivial arguments are valid, and draw
one of several possible conclusions from that. One might conclude that since the
premises are clearly true, we only use innocent statements, thus the questions that
ontology tries to ask are indeed trivially answered:
A The trivial arguments are valid. The questions of ontology are thus easily
answered.
7 Yablo changed his mind about this. A more recent account of his is in Yablo (). His most recent
view is in Yablo ().
i i
i i