NECESSARY SUBSTANCES:

AN EXERCISE IN AXIOMATIC METAPHYSICS

DISSERTATION


A dissertation submitted in partial fulfillment of the
requirements for the Doctor of Philosophy in
College of Arts and Sciences



By
Adam James Labecki
Director: Dr. Brandon Look, Professor of Philosophy
Lexington, Kentucky
2013
Copyright Adam Labecki 2013






ABSTRACT OF DISSERTATION

NECESSARY SUBSTANCES:
AN EXERCISE IN AXIOMATIC METAPHYSICS

My dissertation concerns the restrictions that classical logic, when extended to modality,
and temporality, places upon metaphysics when combined with an ontology of causally powerful
substances. It is an established result of classical quantified modal logic (SQML) that the domain
of discourse is fixed both in terms of what exists at any possible world and in terms of what
exists at any time. That is, SQML entails that everything necessarily exists and nothing could
have existed besides that which exists in fact. Avoiding this outcome would require the
abandonment of classical quantificational theory.
One might wonder if there is much place for contingency in a world of necessary
substances. The occurrences of events and processes may still turn out to be contingent and, I
argue, they provide contingency enough to save the phenomena. In this respect, it is shown that
formal classical logic reproduces the ancient ontological division between being and becoming.
Being consists of a fixed domain of necessary objects (with necessary intrinsic properties),
whatever those objects might turn out to be, while becoming consists of (typically) contingent
events and processes. On this view, Socrates the substance is a necessary being while Socrates the
lifetime is a process that spanned from 469 BC 399 BC. Socrates the substance, however, still
exists as a substance (but perhaps not as a human). Events and processes are understood to be less
fundamental than the fixed substances themselves, since the former are modifications of the
latter. As for the substances, their characterization is mostly negative (i.e., they cannot easily be
identified with any familiar objects) but it is argued that they should be understood as concrete
objects possessing intrinsic natural properties (construed as both qualitative and dispositional)
sufficient to ground the events and processes through which they are manifest and by which we
may fix referents to them. Along the way, the dissertation discusses a variety of topics, such as
the nature of metaphysical modality as well as the nature of metaphysical grounding, and
dispositional properties.


KEYWORDS: Metaphysics, Modality, Ontology, Dispositions, Necessitism






























Students Signiture


Date









NECESSARY SUBSTANCES:
AN EXERCISE IN AXIOMATIC METAPHYSICS



By
Adam James Labecki








Director of Dissertation



Director of Graduate Studies








iv

TABLE OF CONTENTS

Chapter One: Introduction, Overview and Methodology
1.1 Central Objectives and General Overview 1
1.2 Metaphysics and Reality in itself 6
1.3 Metaphysics and the Question of Physicalism 9
1.4 The Conceptual Analysis of Fundamental Reality 12
1.5 Reality in Itself and the View From Nowhere 14
1.6 General Principle before Particular Cases 16
1.7 Metaphysics and the Axiomatic Method 21

Chapter Two: Second-Order Simplest Quantified Modal Logic (SQML
+
) and Possible Worlds
2.1 Sketch of SQML
+
24
2.2 The Domain of Discourse D 28
2.3 Possible Worlds W 29
2.4 S5 and the Metaphysics of Modality 32
2.5 Some General Definitions 33

Chapter Three: SQML and the Threat of Modal Collapse
3.1 The Denial of Modal Collapse (A1) 36
3.2 Necessity and Identity 37
3.3 Existence and Necessity 38
3.3.1 The Necessity of Existence 42
3.3.2 Truth-Makers and the Truth-Making Principle 42
3.3.3 Direct Reference 45
3.3.4 NNE and the Denial of Alien 46
3.4 Contingentism and the Problem of Possibilia 47
3.4 The Barcan Formulas 53
3.5 Some Additional Theorems of SQML
+
58
3.6 Alternatives to SQML 58

Chapter Four: On Previously Proposed Interpretations of SQML
4.1 Exposition of Contingency of Concreteness and Non-concreteness 61
4.2 Critical Evaluation of CCNC 64
4.3 Exposition of Bare Particular Anti-Essentialism 68
4.4 Critical Evaluation of BPAE 71

Chapter Five: Times, Events, and Processes
5.1 Temporal Logic and the Permanence of Existence 74
5.2 The Permanence of Existence and the Denial of Four Dimensionalism 80
5.3 Events and Processes 84
5.4 The Logic of Times, Events, and Processes 86
5.5 Some Event Logic Theorems 91

Chapter Six: Factual Equivalence, Grounding, and Qualitative Properties
6.1 Grounding and Factual Equivalence 96
6.2 Qualitative Properties: X 99
6.3 The Logic of Factual Equivalence and Ground 101
6.4 Some Work for Factual Equivalence, Grounding, and Qualitative 108
6.5 Dispositions 110


v

Chapter Seven: Substance and Power
7.1 The Eleatic Principle and Thick Realism (A2) 119
7.2 Quiddities Leads to Radical Skepticism and Faade Realism 122
7.3 Some Work for Thick Realism 126
7.3.1 The Metaphysics of Dispositions Natural Properties 126
7.3.2 On Concreta and Abstracta 128
7.3.3 New Definitions: Completion of the Core Language 130
7.4 Foundations A3 136
7.4.1 Statement and Some Immediate Implications 136
7.4.2 On the Grounding of Abstracta 138

Chapter Eight: Hardcore Actualism and Branching World-Stages
8.1 General Defense of HCA 142
8.1.1 General Background 142
8.1.2 On Wide Nomological Possibility 144
8.1.3 Against Abundant Metaphysical Possibility 148
8.1.4 Conceivability and Meta-metaphysical Possibility 149
8.2 Hardcore Actualism and Branching Worlds 152
8.2.1 Motivation 152
8.2.2 Branching Stage Logic 152
8.2.3 Branching Histories and the Denial of Presentism 162
8.3 Individual Hardcore Actualism (A4 and A5) 163

Chapter Nine: The Necessity of Fundamental Properties
9.1 The Case against Observable Intrinsic Properties 169
9.2 Problems for Temporary of Contingent Intrinsic Properties? 175
9.2.1 On the Problem of Temporary Intrinsics 175
9.2.2 The Argument from Strong Nominalism 178
9.2.3 A Proof from Unique Modal Equivalence 181
9.3 Relationalism, Presentism, and Temporal Parts 182
9.4 Consequences of A6 183
9.5 The Necessity of Fundamental Properties and Saving the Phenomena 191

Chapter Ten: Objects, Processes, Outstanding Questions, and Concluding Remarks
10.1 Introduction: the Beginning of the End 196
10.2 On the Lifetimes of a Substance 197
10.3 Being and Becoming: The Rationalist versus Empiricist Divide 204
10.4 What Might Play the Fundamental Substance Role? 207
10.5 A Note on Locating Manifest Substances 210

Appendices
Appendix A: Why I am not a Humean 212
Appendix B: Sketch of the Overall Logic 224
References 231
Vita 239




vi


















1

CHAPTER ONE:
INTRODUCTION, OVERVIEW, AND METHODOLOGY

1. Central Objectives and General Overview
This work has sought out to accomplish several key objectives, perhaps too many objectives. Such is a
liability when attempting to do systematic metaphysics. One way to offset this problem is to present a
system around a single theme. Lewis On the Plurality of World (1986a) is centered on Modal Realism
while Armstrongs A World of States of Affairs (1997) is centered on a theory of states of affairs. The
present treatise does not possess single center of gravity, but, like a binary star system, it possesses two
centers of gravity. Just as binary star systems are more frequent than unary star systems, I would like to
suggest that it is more natural for systematic metaphysics to revolve around two central hypotheses, since
the most interesting results arise from the interaction of two distinct intellectual commitments.
The first of these centers is necessitism, the doctrine that everything that possibly exists
necessarily exists. This doctrine has been previously defended by Linksy and Zalta (1994, 1996) and
Williamson (1998, 2010), largely on the grounds that it follows from the simplest quantified modal logic
(SQML). The second center of gravity is substance ontology, a set of doctrines requiring that fundamental
reality consists of substances with powerful qualities and their natural relations.
The addition of substance ontology leads to a considerably different kind of necessitism than that
found in the authors mentioned above. They have attempted to defang necessitism by making
concreteness a contingent property everything necessarily exists but many things could have been
concrete which are not actually concrete while actually concrete things might have been non-concrete.
Ordinary talk of existence and non-existence is replaced by talk of concreteness and non-concreteness.
Necessitism gets to be true without significantly changing how we think about the domain of concrete
reality. I call this doctrine CCNC for the contingency of concreteness and non-concreteness. As we shall
see CCNC turns out to be in violation of substance ontology. Hence the present treatise is effectively
defending the necessity of substantial existence it is a theory of necessary substances. Necessitism
should be regarded as a game changer you cannot accept necessitism and still go home to Kansas. In
this treatise, a fully-fanged version of necessitism is to be reared and unleashed upon the unsuspecting
inhabitants of Commonsenseville.
In terms of methodology and presentation, this treatise is an exercise in systematic axiomatic
metaphysics. Spinozas Ethics made use of the axiomatic method, but it was written before the

2

development of symbolic logic and set theory and hence it does not achieve the full clarity, rigor, and
deductive validity characteristic of geometric presentations. Zaltas Object Theory (1983, 1999) is
presented in a fully rigorous and axiomatic manner, but it is essentially a theory of intensional or abstract
objects, which is typically not, either presently or traditionally, taken to be the central subject matter of
metaphysics. The proper, extra-logical, axioms in this work are aimed at characterizing fundamental
reality, which is understood to consist of concrete substances.
In adopting the axiomatic method, it is necessary to compile a rather expressive formal language.
This language will involve quantified modal logic, quantification over time, events, processes, as well as
a formalization of factual equivalence and grounding operators, which, in turn, permit the formation
of a plethora of defined terms. Some aspects of the formal language are not ultimately employed in
theorem building, but the development of such a language is a matter of interest on its own and may be of
use in future elaborations.
The following format will be used for numbered definitions, doctrines, and theorems:
Df: Definition ..
Do: Doctrine ..
T: Theorem ..
The statement of a doctrine is not an assertion of its truth, since the bald assertion of a doctrine is
precisely an axiom. There will not be a uniform format for axioms. An axiom regarding the formal
language will be prefixed relative to the component of the language for which it is an axiom. For instance
the axioms for propositional logic will be PL1, PL2, and PL3. The six proper, extra-logical, axioms of
the system, will be prefixed with A. I have included a sketch of the entire logic as Appendix B for the
convenience of the reader.
The systematic axiomatic principled-based metaphysics herein presented is intended to provide
an alternative to the neo-Humean metaphysics characterized, among others, by Quine (1950), Lewis
(1986a, 1986b), and Sider (2001, 2011). Neo-Humeanism holds a great deal of sway over most
contemporary debates in metaphysics. I believe that the most attractive feature of neo-Humeanism is the
systematic and elegant development it has enjoyed through the talented efforts of Lewis, along with
further refinements due to Sider and others. If this is right, then the best way to combat Neo-Humeanism
is to show how systematic metaphysics can be done while denying the central tenants of Humeanism.
(Although, I must admit that system presented herein could use a little more elegance in places). In my
judgment, there is no proper name relative to which I might label my cause, but it might well be
described as form of Neo-Rationalism, if the use of the axiomatic method did not already give this away.

3

I initially intended to include a full frontal assault upon Lewis doctrine of Humean
Supervenience within this first chapter, but doing so would have made the chapter too long and so I
decided to include much of it in Appendix A. I also decided that the positive case for my approach is
more important than criticizing alternatives. Nevertheless, the case against Humean Supervenience does
strengthen the methodological argument of the present chapter, since it provides additional reason to drop
the conceptual analysis of platitudes to and take up theorizing directly concerned with the nature of
Reality in itself. But to summarize, Appendix Two proceeds in two phases. First, it is argued that
Humean Supervenience is incompatible with every mainstream interpretation of quantum mechanics due
to the fact that it presupposes and requires classical physics. Second, it is argued that, at bottom, the
whole business of Humean Supervenience (and this goes for most location problem metaphysics) is
really a red herring that draws attention away from the unavoidable fact that, according to full-blooded
Humeanism, everything is, fundamentally speaking, totally chaotic. Things just happen and that is all
there is to it.
The present chapter will be largely concerned with the matter of methodology and
metametaphysics. After this overview, it begins with the idea that the primary task of metaphysics
concerns the nature of Reality in itself, or being qua being, and ultimately concludes that investigations
pertaining to Reality in itself are best conducted by employing the axiomatic method.
Chapter Two presents the axiomization and semantics of second-order quantified modal logic
(SQML
+
). Some discussion is included on the general nature of modal metaphysics and the nature of
possible worlds. I tentatively adopt a state description form of possible world ersatzism, which view
(or rather, something like it) receives more systematic development in Chapter Eight.
Chapter Three presents some basic theorems of SQML while discussing their implications. The
first proper axiom A1 consists in the assumption that there is no modal collapse (i.e., things in general
could have been otherwise than they are). Since it is impossible for different objects to have existed, it
must be that things could have been quite otherwise than they are, at least with respect to their manifest
appearance. In particular, manifest sortals such as wolf and oak cannot be essential properties (that is,
if any thing really is a wolf or oak in the first place).
Chapter Four considers and criticizes two varieties of necessitism. The first variety is
championed by Linsky and Zalta (1994, 1996) and Williamson (1998, 2010). As noted above, these
authors attempt to defang necessitism and other surprising SQML theorems by introducing a new
ontological category: the contingently non-concrete. The theory is criticized on several fronts, including
the fact that it introduces vast modal differences between objects that do not differ with respect to their
categorical (or qualitative) properties.

4

A second interpretation of SQML, which makes use of bare particulars and radical anti-
essentialism, is also discussed. It is found to be a step in the right direction but is rejected on the basis of
its commitment to bare particulars, a deflationary attitude toward modality, and incompatibility with
substance ontology.
Chapter Five introduces temporality into the picture. It is shown that objects are not only
necessary but also permanent (although, given necessitism, this might not come as a surprise). This,
combined with the permanence of identity, leads to the elimination of four-dimensionalism as a viable
contender, since it requires temporarily existent objects within the domain of discourse.
The second part of the chapter introduces a robust logic of fine-grained events (i.e., simple events
analyzed as first-order formulas with an added time relata). Given both singular and plural quantification
over simple events, an expressive range of definitions can be produced, permitting compound events,
various kinds of processes, and terms analyzing these items.
While events are necessary existents, they do not necessarily occur. This is because they are like
propositions in that quantification over them is into formula position rather than term position. Just as a
proposition can exist without being true, an event can be said to exist without occurring.
Chapter Six introduces factual equivalence, grounding, dispositions, as well as the only primitive
term in the language that is not explicitly analyzed in the theorys model semantics, namely that of a
qualitative property. That there should be at least one such primitive is not surprising concrete reality
cannot be explained wholly in terms of set theory. That there should be exactly one such primitive is
optimal for obvious reasons of conceptual parsimony.
Factual equivalence is effectively my version of supervenience or grounding, at least in cases in
which that which supervenes or is grounded is taken to be nothing over and above the base or ground.
In such cases two items (or classes of items) are conceptually different but the same in reality. The one
might make use of more natural, objective, or precise, vocabulary than the other, but both refer to the
same part or mode of Reality. What I mean by grounding, as opposed to factual equivalence, concerns
cases in which the grounded is something over and above its ground (i.e., non-reductive grounding). The
primary example of grounding would be causation, although it would also include any cases of
emergence. Both factual equivalence and grounding are axiomatized and formally modeled. Dispositional
predicates, in the form of something like habituals, are defined and analyzed in terms of modality,
events, and partial grounding, where dispositional properties are properties assigned to sets of such
ascriptions. A dispositional property is that in virtue of which certain dispositional ascriptions may be
said of an object.

5

Chapter Seven introduces the first two truly substantive axioms: A2-A3. A2 requires that each
dispositional property is identical with some categorical property. Natural properties are then defined as
those which are both dispositional and categorical. The introduction of natural permits a variety of new
definitions, including concrete, abstract, fundamental and natural relation. A3 requires that all
truths (with slight qualification) should be factually equivalent to some plurality of fundamental events
(events only involving perfectly natural properties or relations). This will require that the whole of
modality is to be explained in terms of actual fundamental concrete reality hardcore actualism (HCA).
Likewise for abstracta, provided that there are any, must also be grounded in the concrete fundamental
base.
Chapter Eight is mostly dedicated to defending and elaborating HCA. It is argued that HCA can
deliver a robust-enough modality for the purposes of modal metaphysics and that possible worlds are best
analyzed as possible histories of the actual world. The remainder of the chapter introduces two axioms
related to HCA, A4-A5, as applied to individual objects, on the basis of which concreteness and
abstractness turn out to be necessary properties.
Chapter Nine considers arguments in favor of the fixity (i.e., permanence and necessity) of
intrinsic properties. It is untimely found that standard arguments to this effect are wanting but that there
are some good arguments in favorer of the fixity of what I call strongly intrinsic properties, which
would include those which are perfectly natural or fundamental. On this basis it is shown that all
contingent properties are to be explained in terms of contingent natural relations between fundamental
objects and so both priority monism and existence monism turn out to entail modal collapse there must
be a proper plurality of fundamental objects.
Chapter Ten concludes by tying up a few lose ends and suggesting some lines of further
investigation. As far as the relation between Reality in itself and the manifest image is concerned, the
former may be conceptualized as the realm of Being, necessary existents that never change in their
fundamental nature while the latter may be conceptualized as the realm of Becoming, namely events and
processes abstracted from the fundamental objects that underlie them. Ordinary names for objects of both
the manifest image and natural science refer, at most, to instances of a special sort of process involving a
necessary object, which processes I call lifetimes.
Hence, we come upon a partition between Being and Becoming, where the latter is the expression
or manifestation of the former and the former is the underlying Reality of the latter. Being is reality as
something that can be thought while becoming is reality as something that can be experienced. Platonists
and Rationalists have historically testified to the former, to the dismay of Empiricists who deny that we
can be justified in believing anything above and beyond that latter.

6

Some additional speculations are considered regarding how the fundamental base might be
further related to the manifest image. In particular, some prospects for composite objects and emergent
properties are discussed.
2. Metaphysics and Reality in Itself
I take the claim that that metaphysics at least involves theorizing about Reality in itself to be a truth of
convention it is part of what people in the profession mean when they say metaphysics. In his
editorial introduction to the collection Metametaphysics, David Manley writes the following:
Most contemporary metaphysicians think of themselves, not primarily concerned with
representations of language and thoughts, but with the reality that is represented (Manley 2009: 3).
Gone, it is said, are the days when metaphysics was the handmaiden of ordinary language philosophy or
conceptual analysis. Mainstream ontological realists such as Lewis, Fine, Armstrong, van Inwagen, Heil,
Schaffer, Lowe, and Cameron regard metaphysics as a discipline that is interested in discerning and
explicating the objective and fundamental joints of reality. Regarding the cutting of nature at its joints,
Sider writes: It means inquiring into how the world fundamentally is, as opposed to how we ordinarily
speak of it (Sider 2011: 1).
Contemporary metaphysics, however, is not simply about Reality in itself. For instance, consider
the following two passages from Fine (2001):
The conception of reality we are after is simply the conception of Reality as it is in itself (Fine
2001: 25).
The aim of realist metaphysics is to render the world intelligible in terms of the distinction
between what is real and what is not real; and its task is complete once it becomes clear how what
is apparent, or not real, is to be rendered intelligible in terms of what is real. (Fine 2001: 29).
Fine is interested not only in the Real but the way in which the apparent is explained by the Real. He is
here alluding to what is also known as the general location problem (Frankfurt 1998): how do we locate
the non-real or apparent within the domain of the fundamental? How does our subjective view of the
world map onto our objective view of the world (Nagel 1989)?
Historically speaking, interest in location problems is probably, in part, due to the fact that
analytic metaphysics emerged during a time when the philosophy of language was at its height and logical
positivism was still a recent memory. Even after metaphysical realism came into full swing, philosophers
like Lewis (1986b) continued to believe that it is important to keep our theories grounded in common
sense, including a wide range of platitudes, some regarding matters of general principles, others
regarding beliefs about particular matters of fact. The platitudes of our ordinary beliefs and language are

7

to be explained in terms of what is fundamentally real, which is typically thought to be the domain of
micro-physics.
The present treatise is primarily concerned with developing a theory of Reality in itself, where
limited attention will be given to explanations of the non-fundamental. A theory of reality must, of
course, save the appearances, but it need not weigh in on whether there are musical works or tables and
chairs. It is not so much that such questions are not important; it is rather that developing a theory of
Reality in itself is primary both in the temporal sense (i.e., we cannot do the location problem until we
know what we mean by Reality) and in the essential sense, in that such is central task of metaphysics. Or
so I claim.
In this respect I stand with Aristotle, who described metaphysics as the study of being qua being
1
.
Contemporary philosophers tend to assert ontology is concerned withwhat there is while metaphysics is
concerned with what it is like. For Aristotle, however, metaphysics is concerned with what is being as
such like, where being as such is understood to be primary substance or substantial essence.
2
What, in
general, must something be like if it is to possess the being of substance (i.e., if it is a being in the primary
sense of the term)? The study of being qua being has traditionally been called metaphysica generalis. It is
the study of the general features that characterize Reality in itself. It is not precisely about what there is or
what it is like but about what fundamental reality is like or, more precisely, it is about what something
must be like if it is to be deemed fundamental.
Location problem metaphysics is the result of several tendencies within analytic metaphysics. I
have already mentioned its birth within a philosophical environment in which the linguistic turn was still
strong. Anything that was going to grow out of that environment was going to have to resemble linguistic
philosophy in some non-trivial way, and so a conceptual turn (where our concepts can be gauged on the
basis of the analysis of ordinary language) was natural.
That is not all. There are additional reasons, not least of which is the presumption that Reality in
itself is effectively the domain of micro-physics. Typically, nothing more is said of Reality in itself
beyond some classical toy model theory which, being classical, does not even have much to do with
anything studied by contemporary physicists. Moreover, metaphysics was permitted to return from exile
only by taking on a Humean form, under the leadership of Quine. The methodology differs from orthodox
Humeanism in that perceptions are replaced with objects which are thought to be posited by micro-

1
Although he also called it study of the causes of being or the study of the highest being. The first two being
largely equivalent and concerns something like the study of the general features of being while the third concerns
philosophical theology.
2
The former is the doctrine of the Categories while the latter appears to be the doctrine of the Metaphysics. It is,
however, far from clear just what Aristotle meant by substantial essence.

8

physics, typically sets or pluralities of spacetime points. Hence, what was anti-metaphysics in Hume
has morphed into contemporary realists metaphysics. Contemporary metaphysis, however, remains anti-
metaphysical in the traditional sense of the term in that it does generally permit metaphysics to develop
speculative theories about the nature of Reality in itself. As such, metaphysicians typically have relatively
little to say about fundamental Reality as it is independently of our subject point of view. But now it
seems as though we are not directly interested in Reality in itself so much as how Reality as it is for us
maps onto an underdeveloped conception of Reality in itself.
It is typically thought that the non-fundamental supervenes upon the fundamental, or is
realized by it, or is grounded in it, etc. In each of these cases, however, it is usually said that the
supervening, realized, or grounded item is nothing over and above its location in the fundamental base.
To my mind, if the Bs are reducible to the As and if the Bs are nothing over and above the As, that can
only mean that the Bs just are the As, subjectively (or pragmatically, or conventionally etc.) conceived.
That is, the difference between the As and the Bs is conceptual, or a difference in the understanding, not a
difference in Reality, because at most the As belong to Reality. The difference between fundamental
reality and non-fundamental reality is nothing more than the difference between the view from nowhere
and the view from wherever one empirically finds oneself.
I believe that location problems, or functionalism in general, has retained a strong presence
among analytic metaphysicians for several reasons. First, as mentioned above, most metaphysicians
delegate nearly everything about Reality to physics. As such, metaphysicians cannot find much work to
do beyond attempting to link what they take to be the scientific image with the manifest image. Second,
there appears to be concern as to how we find ourselves and the manifest image in the scientific image.
I do not deny that this line of questioning is philosophical but I do deny that it is particularly
metaphysical. Indeed, it seems to lie in the intersection of M&E and value theory, since its purpose seems
to be more therapeutic than theoretical. Third, very often mainstream metaphysics seems to be locked
into ideological battles between naturalist/physicalists and non-naturalists/physicalists, many, but not all,
of whom are theists. Reductive analysis is here motivated by the desire to show that we need not posit
anything beyond what is studied in the natural sciences everything that needs to be explained can be
physically explained. Metaphysics has had a long history of playing bondmaid for one ideology or
another and I believe that such trappings undermine any pretension it may have to concern itself with
Reality in itself. Ideological commitment breeds partisanism and dogmatism, the twin evils of free
intellectual inquiry, since the questions that should be asked, and the right answers to most of the
interesting ones, are decided in advance of honest intellectual inquiry.

9

It does seem, however, that, within mainstream metaphysics, there has been a gradual drift away
from conceptual analysis and toward the development of theories of Reality in itself. I would like to give
this tendency a quickening nudge. To this end I shall argue that we would do better if we did not always
assume, from the beginning, the domain of Reality in itself is co-extensive with the domain of
fundamental physics. This will free metaphysical investigations to develop full-blooded theories of
fundamental Reality.
3. Metaphysics and the Question of Physicalism
Fundamental reality and physical reality are conceptually distinct. We have certain beliefs about what
it takes for something to be fundamental that are quite independent of anything that has been learned by
way of physics. For instance, it is manifestly known by the natural light that fundamental reality must
be independent (if it were dependent how could it be fundamental?)We do not know in advance that
whatever it is that physics is ultimately studying (i.e., whatever the nature of the physical turns out to be)
will be the sort of thing that even could be fundamental.
The reader might object that I have begun to sneak conceptual analysis back into the picture,
having just set it aside. Conceptual analysis will continue to play an important role but it is conceptual
analysis about Reality in itself that is in question, not conceptual analysis about ordinary common sense
postulates. I shall have more to say on this below.
I shall argue in that there is no need to assume in advance that physical objects are the sort of
things that could play the role of fundamental objects role. First, our best physical theories are not
theories of fundamental physics, much less of fundamental reality simpliciter. Second, there are a wide
number of speculative fundamental theories that may eventually be recognized as our best physical
theory. Given that we do not know what fundamental physical reality is really like, our confidence that it
must be fundamental must be limited.
It is well known that the role of best theory of physics is jointly shared by General Relativity
(GR) and Quantum Mechanics (QM). These two theories effectively share the best theory role because
neither one by itself can explain all of the phenomena and, in this sense, each is incomplete. This, in itself,
would not pose a problem were it not the case that GR and QM cannot be consistently combined and
hence at least of one the two is not merely incomplete but false.
3
GR, for instance, predicts the existence
of physical singularities of 0-dimension while QM requires that any physical object should have a
wavelength of some positive length. Moreover, almost every quantum theory of gravity will require that

3
Strictly speaking, only sentences or propositions can be true or false, but we might regard a theory as a conjunction
of propositions (its axioms, for instance), where the theory is false just in case the conjunction of its axioms is false.

10

spacetime be quantized (inasmuch as fundamental spacetime is retained at all). For the same reason, on
QM it is inadmissible to treat individual spacetime points as physical objects and predicating physical
properties to spacetime points leads to physically meaningless infinities.
4

Accordingly, GR is generally regarded as an effective theory. There appear to be two distinct uses
of the term effective theory. First, a theory is said to be effective if it is adequate in predicting and
describing a certain range of observations but it is denied that the items of theorys object-language
correspond to actual entities or mechanisms in nature. Second, a theory is said to be effective if its subject
of study is the effect of the subject matter of some other theory (possible or actual). Of course, if there are
no genuine levels of reality, the two accounts of effective will turn out to be equivalent. GR is an
effective theory in both senses of the term and, accordingly, it is not a theory to which one should turn
when deciding what there is. It is effective in the first sense due to its inconsistency with features of QM
that will be inherited by all future physical theories. Both GR and QM are thought to be effective in the
second sense, since both are expected to be superseded by some unified theory of physics (i.e. quantum
gravity).
The upshot is that our best physical theories are not theories of fundamental physical reality and
hence are not theories of fundamental reality simpliciter. Despite these considerations, working
metaphysicians tend to proceed as if Special Relativity (which was already superseded by GR) was our
best theory of fundamental reality (and this is when they even bother to go beyond ancient atomism).
Hence we are often told that fundamental reality consists of something like spacetime points (typically
formalized in Minkowski spacetime), but theoretical physicists are increasingly coming to regard the
spacetime of SR or GR in phenomenological or emergent terms.
5

Much intellectual effort has gone into the production of unified theories, nearly all of which are
theories of quantum gravity which pertain largely to the physics of the Plank-scale (i.e., high energy
physics). The most studied such theory is Membrane theory (M-theory), which is effectively the successor
of string theory (although the name string theory has retained its place within popular parlance).
Strings are effectively 1-dimensional branes (or 1-branes) while 2-branes are sheets. According to M-
Theory all properties of fundamental particles or fields are understood in terms of brane-vibrations. There
are, however, other quantum gravity contenders, including Loop Quantum Gravity (LQG)
6
and Causal

4
This is partly because on QM, for nearly any determinable property F, the set of F-determinants will have a
countably cardinality (i.e., the magnitudes of properties do not take Real values).
5
The physicist sense of emergent is typically ambiguous between merely phenomenal and ontologically
emergent.
6
For a relatively involved introduction to LQG see: Rovelli (1998).

11

Sets
7
(both of which are background independent theories which generate quantized space and quantized
spacetime respectively). A notorious difficulty for theories of quantum gravity is that, by dent of being
theories of high-energy physics, any novel predictions made by such a theory typically require energy
levels which are many magnitudes greater than those which can be produced by our most powerful
particle accelerators.
8
Accordingly, empirical methods cannot seem to decide between the various
contenders and there is little reason to believe that this will change in the near future. Moreover, many of
these theories continue to be undeveloped in various respects and continue to face significant
computational challenges. This is why GR and QM have retained their best theory status despite their
known deficiencies.
One might have expected that metaphysicians would respond to this situation by engaging in
nuanced debates within theoretical physics. Hence metaphysicians might have asked which interpretation
of QM we should adopt and, perhaps more importantly, which approach to unification is most promising.
Indeed, many physicists consider such theorizing to be closer to philosophy than science: so why should it
be of such little concern to metaphysicians? Such questions are of concern to philosophers of physics.
Indeed, if the subject matter of metaphysics is Reality in itself and if fundamental reality is wholly
physical, metaphysics should be equated with the philosophy of physics (perhaps extended to the
philosophy of chemistry and biology).
There is certainly important work to be done here. When reading what physicists have to say
about the ontological and metaphysical interpretations of their own theories, one comes to value the
conceptual and analytic skills that are instilled by an education in analytic philosophy. Distinct concepts
are conflated while others are used without any clear meaning. Insufficient inferential caution leads to
frequent speculative jumps in reasoning. There is much philosophical work to be done here and, on
physicalist assumptions, metaphysics would be the study of what Reality must be like in light of present
and anticipated physics. Since physics itself is divided in many respects, such theories will need to be
content with features common to most contenders or develop theories which are conditional upon the
eventual success of a given contender. Physics as such would not be understood to provide us with the
fundamental domain, but would provide the sole datum on the basis of which one theorizes about the
fundamental domain and what it is like.

7
For an introduction to Causal Sets by the theories founder, see Sorkin (1990). The basic slogan behind causal sets
is that Order + Number = Geometry.
8
Some theories make predictions which could be detected in nature, such as on the surface of an event horizon, but
since such events would be microscopic, measurement devices would have to be brought very close to such regions
and survive long enough to make take the required measurements. Moreover, at present we have no means to even
travel to any black holes.

12

Ladyman and Ross (2007) and Ney (2012) advocate a Neo-Positivists approach to metaphysics
which is, in some ways, similar to the stance outlined above.
9
Neo-positivists differ among themselves on
many details of this approach, but, roughly speaking, the primary theoretical constraint on metaphysics
should come from the natural sciences and metaphysical theories should be restricted to problems that are
motivated by the natural sciences. In particular, the concern with reconciling the manifest image with the
scientific image will be largely dropped or, at least postponed until a sophisticated and scientifically
motivated theory of Reality has been developed.
Neo-Positivism, broadly construed, is a plausible approach to metaphysics and, indeed, it is
already effectively an active and exciting research program, since much of the current work in the
philosophy of physics could be regarded as Neo-Positivist metaphysics.
I do not believe that metaphysics as a whole should be identified with the philosophy of physics.
My main line of contention is not that that the philosophy of physics is too narrow in its scope (even if
one were to integrate chemistry and biology) but that it does not seem obvious that fundamental reality
is co-extensive with physical reality. There is no agreed upon way to even define physical, since we
are not entirely sure what sort of entities will eventually be posited by physics (indeed, it is difficult to get
clear about some of those entities which have already been so posited). This is especially so given the fact
that we presently have no comprehensive theory of fundamental physical reality. Since we do not know
what the fundamental physical entities will be like, we do not know whether or not they will be the sort of
things that could be regarded as fundamental. It is readily admitted that physicalism is a plausible enough
hypothesis and so, when presenting a metaphysical theory, one is justified in assuming its truth from the
beginning. What is not justified, however, is that physicalism is obviously true and hence should be a
given in any serious metaphysical theory.
4. The Conceptual Analysis of Fundamental Reality
Metaphysics will require some conceptual analysis at its core in that we must have some conception of
Reality or primary substance. There are, of course, rival conceptions of substance, but among these there
are some features that are especially common. I already mentioned that independence is, and really
must be, ubiquitous. A substance is not sustained in being by something distinct from itself. Given
independence, a substance should also possess at least some intrinsic properties, being those properties it
possesses independently of the states of other substances and its relations with said substances. Such are
properties that it has only in virtue of itself (including its parts). That is, substances cannot be wholly

9
Although Ladyman and Ross would be less comfortable with the position just described than Ney. The former
relegate the function of metaphysics to the finding of structural similarities between the various sciences (including
the social science); the interpretation of physical science is secondary.

13

relational. As far as I am aware, all mainstream realist metaphysicians are in agreement on these two
points, although not all of them would use the term substance to describe their fundamental entities. I
shall later have occasion to introduce some additional requirements which are somewhat contentious,
mostly notably that substances must be causally powerful.
It is worth nothing that some Neo-Positivists, such as Ladyman and Ross (2007) conclude that the
objects of science in general and physics in particular are wholly relational. There are no individuals with
intrinsic properties, there are just relational features of physical reality. Their argument proceeds along
two lines. First, it is argued that scientific realism about the individual entities posited in physics theories
is highly implausible (i.e., current individual posits will most likely play no role in future best theories)
but that central structural elements are preserved between successive theories. Second, it is argued that
quantum physics is best interpreted as a theory of structures rather than of individual particles. I shall not
evaluate these arguments here, except to say that the first enjoys prima facie plausibility while the second
requires, in the very least, a prohibition on metaphysical postulates that are not made in physics and
which are not required for a realist interpretation of physics. One might, however, wonder whether or not
realism requires that physical relations be grounded, in part, by relata bearing at least some intrinsic
properties. This is bound to be a point of contention and perhaps primitive disagreement.
At any rate, relationalists like Ladyman and Ross refuse to acknowledge any a priori claim that
physical Reality in itself cannot be wholly relational. Nearly all mainstream metaphysicians disagree and,
given that the great majority are physicalists (or something close enough), nearly all insist that physical
objects must have intrinsic properties. Heil (2012) makes a moderate demand in this regard, in that he
suggests that, if Ladyman and Ross (2007) are right about physics, then one would opt for existence
monism, being the doctrine that the universe as a whole is the only physical object and that it has all of its
natural properties intrinsically.
As for myself, I do not assume that the fundamental domain is physical, but I do insist that its
members should have intrinsic properties. I simply cannot conceive how reality could be absolutely
relational but I can conceive how physical reality could be wholly relational provided that it is non-
fundamental. This could, of course, be my own shortcoming, but I do suspect that a wholly relational
account of the world provides hardly any more sense of what the world is like than instrumentalism and
hence non-realism. At any rate, I am unable to embrace this view, in the very least, because I am not even
sure what it really means to say that it is structure all the way down. I am not sure how such a reality
could even be concrete as opposed to abstract (wholly structural actually might be a good way to define
abstract). I need not, however, insist that physical reality involves a plurality objects with intrinsic
properties. Indeed, if physical objects were wholly relational, then I would have to suppose that they

14

possessed some sort of shadowy and dependent existence, as Plato might have suggested in the allegory
of the Cave.
That is, if physics can be understood without recourse to individuals with intrinsic properties and
if fundamental objects must have intrinsic properties, we would have further grounds to deny the
obviousness of physicalism. It would be the case that the domain of physics is presented in such a way
that it cannot play the role of fundamental objects.
5. Reality in Itself and the View from Nowhere
Given that that the primary task of metaphysics is to discern Reality as it is in itself, I suggest that when
we do metaphysics we strive to assume what Nagel (1989) calls the view from nowhere. When we take
up the view from nowhere we no longer regard average conditions in the lower atmosphere of the planet
we call Earth to be normal, because they are certainly not normal. We no longer entertain causation by
absence or regard the striking of the match as the cause of the fire. This is because we are no longer
employing casual notions to improve our capacity to control the world or avoid certain undesirable
outcomes or to assign responsibility (i.e., we do not distinguish between the salient and non-salient
contributions to a phenomena). In order to do metaphysics we must remove ourselves as far as possible
from ordinary daily concerns and purposes, from ordinary language, and from common opinions,
otherwise we will fail to take up the view from nowhere.
Is it even possible to take up the view from nowhere? Is it not an attempt in futility to attempt to
look at the world from no perspective whatsoever? How could we even dream of knowing things as they
are in themselves, when we, the knowers, can only know things relative to our cognitive capacities?
It will not be possible for me to adequately address the concerns expressed in such questions.
These questions are effectively asking how is metaphysics even possible? The entire project taken up in
this treatise, of course, presupposes that metaphysics is possible. I shall, however, say a couple of things
that might alleviate this concern, at least a little. First, I accept that most of what we think we know about
the world has little to do with the way it is in itself. Second, I accept that what we can know about
fundamental reality is probably quite limited and in many respects negative. In the end, the particular
points of disagreement between myself and the metaphysical conventionalist (pragmatist, skeptic, etc.)
might turn out to be relatively few in number, but they will certainly substantive in nature.
We should also avoid taking the view from nowhere too literally. Perfect objectivity is a goal
that we might never wholly attain but we can make partial success. We begin by understanding multiple
subjective views rather than just our own. The road to objectivity begins with more subjectivity, but not

15

more of my subjectivity. At first we take up the views of various different contemporaries nearby; then
we take up views from those who do not live near us or who lived during past eras. All the while we
abstain from understanding the less familiar in terms of what is familiar and local to our own self-
centered view.
We then proceed to zoom out from human perspectives by the aid of experimental science,
considering the very large and the very small, the very hot and the very cold, exposing ourselves to the
very alien behavior of each domain, whereby we further loosen our sensibilities from those which we
inherited from our phylogenetic lineage. The fact that we are within the Milky Way does not prevent us
from deducing what it looks like from outside. Indeed, if the universe were finite it would be possible for
us to know what it looks like from outside (any map of the whole universe requiring such a view), despite
the fact that there is no outside. We can accurately see things from points of view that do not even exist.
From there, we move onto the formal sciences, mathematics and formal logic, where the senses
are suspended and we have nothing but thought to guide us. We find that we are able to think about
objects that can in no way be constructed in a step-wise process (not even by God) and about which no
mental intuition provides an adequate representation.
10
It is among these inhuman artic heights that
something close to the view from nowhere may be attained.
As such, the view from nowhere is not strictly a view that is supposed to be somehow
completely non-subjective. Every view is a form of subjectivity. The view from nowhere is a form of
subjectivity that possesses a disinterested interest in knowing things as they are in themselves. Relative to
Platos allegory of the Cave it is the stance that one comes to take when turning from the shadow play on
the caves back wall to the sunlight gleaming from the tunnel leading to the world above. Indeed, if one
pleases, one might replace the the view from nowhere with the view from far above.
If we are to put location questions regarding the manifest image and other anthropological
concerns to the side in order, first, to develop a theory of Reality in itself and if this requires taking up the
view from nowhere, then we should also put aside just about everything we think we know by way of the
manifest image. We shall assume that ordinary everyday knowledge only pertains to appearances or
shadow plays unless we have specific reason to think otherwise. To be sure, in the end these matters
should be explicable in terms of Reality, at least insofar as they are mere appearances, but before we
worry about that we must first to turn away and forget about them for a while. Dont worry they will
still be there when you turn around again.

10
Those empirically-minded mathematicians and logicians who lose their nerve at this point are known
constructivists and intuitionists respectively.

16

How then are we to proceed? Let us take stock of where we are. We have decided to put to the
side all of our platitudes or ordinary beliefs about particular cases. It seems that we have no choice, then,
but proceed by way of general principles.
The linguistic turn in general and ordinary language philosophy in particular privileged particular
cases over general principles. Proceeding in a Humean manner, particular cases constitute the given while
general principles are merely there to lend organization and unity to our understanding of the manifold of
given facts. On this view, principles have no rights against cases. When there is a mismatch between the
two, principles must, in general, change to fit the cases. Given that language and meaning is a wholly
conventional affair this is reasonable. Ordinary language is in no need of correction by theory. In the
case of realist metaphysics, however, we cannot take our ordinary beliefs as legitimate data. Moreover, it
should be assumed that we cannot do too good a job at escaping the trappings of implicit pragmatic and
conventionalist interest, and so a healthy dose of skepticism is required in order to free ourselves from
these trappings.
6. General Principles before Particular Cases: Neo Rationalists Metaphysics
In what way are we to do realist metaphysics in a principled manner? On what basis can an investigation
into Reality in itself proceed and on what basis can we keep metaphysical speculation responsibly
constrained?
The metaphysics of conceptual analysis, with its location problems, subjected itself to the
constraints of ordinary language and a fairly wide variety of common sense platitudes, where many of the
latter can be discerned by analysis of the former. These constraints, while relevant in the context in which
they were initially employed, are not terribly relevant for the study of Reality in itself, since the language
and postulates largely concern items of the manifest image, or the non-real, as Fine puts it. What new
constraints might we put in their place? I shall propose three kinds of theoretical constraints or
touchstones: common notions, formal logic, and actual physics.
Where conceptual analysis draws upon a wide range of our pre-theoretical beliefs as datum,
realist metaphysics should restricts itself to what the Medievals called common notions, which the
Rationalists called innate ideas. In particular, and as mentioned above, we should limit ourselves to
platitudes that pertain to what it means for something to be a fundamental object or substance.
Where conceptual analysis draws upon ordinary language and linguistics in the identification and
clarification of our platitudes, realist metaphysics should draw upon formal classical logic, for the

17

clarification and analysis of common notions (although some common notions will require extra-logical
clarification).
Finally, where conceptual analysis draws upon classical physics for the construction of toy
models of fundamental reality, realist metaphysics should draw upon non-classical physics (i.e., the
physics studies by physicists) in order to (1) discern whether physical objects might play the real
substance role, (2) what Reality is like if physical objects do play the real substance role, and (3) as a
source of intellectual inspiration (i.e., there might be some good ideas here that we can steal and modify
for our own purposes).
Realists metaphysicians adopting these constraints will not be of the Neo-Positivist variety. It
might be described as Neo-Aristotelian but, as mentioned above, the specific path that I shall be charting
in this treatise is probably better described as Neo-Rationalist.
Restricting initial common sense postulates to those which might be considered to be common
notions is, again, motivated by the fact that we only want to begin with those postulates which might
plausibly apply to Reality as it is in itself.
Phenomenologically speaking, common notions are conceived as if through the natural light, or
Platonic recollection, or as if they were innate to the human mind. Aristotles ten categories and Kants
Categories of the Pure Understanding are two exemplary attempts to codify a complete list of common
notions regarding our conception of primary substance, or the idea of an object in general.
I shall abstain from making serious appeals to the natural light or recollection, since such claims
would be unnecessarily bold and are otherwise unconvincing to anyone who is not already a devoted
member of my choir. The question of innateness is also somewhat controversial, but not terribly so if one
has a broadly Kantian or Chomskyan view of innateness in view.
Even where innateness is accepted, it does not necessarily tell us anything beyond how we are
mentally compelled or constrained to think about Reality and so the innateness of these concepts does
little to support the claim that they are reliable. In other words, the existence of innate ideas does not
entail that there is innate knowledge about the world. Kant did not have to worry about this since he did
not believe that the Categories apply to Reality in itself, while Chomsky is only concerned with human
languages. How is the transition from innate ideas to innate knowledge possible? Descartes tried to move
from innate ideas to innate knowledge by way of a benevolent God. Are there any less contentious ways
in which our innate ideas pertaining to Reality might be deemed reliable? One might suppose that the fact
that we manage to find our way around the world is evidence that we know something about it, on the
basis that false beliefs are less adaptive than true beliefs. While this may be true relative to middle-sized,

18

low-energy objects within our Earthly environment, it does not follow that our basic intuitions are reliable
when making judgments about fundamental reality. As soon as we suspend matters that are only of
pragmatic importance to us, phylogenetically inherited cognitive features need not be of any relevance.
11

My response to these doubts is as follows. We do not really know whether or not metaphysics is
even possible and I am doubtful that this matter can be decisively resolved in any manner that would lend
full justification to the pursuit of realist metaphysics. As mentioned above, the project around which this
treatise is organized presupposes that realist metaphysics is possible. Hence, what I am claiming is that, if
metaphysics is possible then our common notions pertaining to Reality provide a reliable guide to the
fundamental nature of Reality.
Just as Chomsky argues that the range of possible grammars is relatively narrow, my hypothesis
here is that there are very few distinct conceptions of Reality that are intelligible to the human
understanding. I shall leave the reader who is at least moderately familiar with the history of philosophy,
be the judge of the plausibility of this claim.
In particular, I claim that relationalism does not pass the test. What about those who claim to find
relationalism to be perfectly coherent? First, I believe that many relationalists will readily admit that
relationalism goes against the grain regarding how we are hardwired to think. The relationalist will not,
of course, concede that relationalism is cognitively impossible. This is where I am forced to pull a van
Inwagen and flatly state that I do not believe that anyone with a robust sense of reality could really
believe that fundamental reality is wholly relational. Perhaps someday I will free myself of my current
intellectual habits and resend this statement, but for now I can only consider those possibilities that I find
conceivable.
In various places I will have occasion to explain why I think that Humeanism, with its denial of
genuine causation, also fails to pass the test. This claim is, of course, contentious. Most of my rivals will
agree with me about relationalism but many will be shocked by my claim that one cannot be both a
Humean and a realist (although it is worth noting that Hume himself was not a realist).
If I am right some of our common notions will factor into any possible theory of fundamental
reality. It is in this vein that Heil (2012) writes the following:
Fundamental ontology is in the business of telling us what the universe must be like if any theory
is true. (Heil 2012: 2)

11
Kant went even further. From the fact that we cannot help but to think about the world as if p were true of it, he
concluded that p is not true of the world.

19

The conditional nature of the statement makes this a claim that even skeptics about objective knowledge
might accept. To be sure, it might be the case that human cognition, at least in its present state, is simply
incapable of producing or understanding a true theory of Reality in itself. But if we are able to produce
such theories then some of our common notions must be true (and Heil happens to be a fellow substance
ontologist in this regard).
It should be noted that Hume rejected the Aristotelian tradition by following through on his
radical empiricism, which inhibits attributing such common notions upon reality. Hume was also
proceeding in a principled manner, only his is a principled empiricism rather than a principled
rationalism. Aside from the core principle of empiricism, however, his philosophy turns out to be
particularistic. This is because the content of experience is always and everywhere wholly particular.
Strict empiricism is general and principled at its root but everywhere else it is wholly particular between
the trunk and the leaves there are no branches. This is why, according to Hume, we must rely so heavily
on custom. Neo-Humeans, with their appeal to particular platitudes follow their master in this respect but
they depart from their master in that they believe it is possible to do substantive metaphysics.
The second set of principles and constraints consists in the injunction, in the spirit of Parmenides,
Plato, and Spinoza, to follow the logos where it may lead. I shall attempt to reap maximal benefits from
this constraint as follows.
First, I shall maintain strict adherence to classical propositional and quantificational logic. Why
the strict adherence to classical logic? When seeking after general principles and constraints classical
logical axioms present the most obvious starting points. In this sense my approach is theoretically
conservative, but my conservatism sides with general principles rather than particular cases. We are not
going to adjust our principles in order to get the results that seem right to us.
There is also some reason to believe that the axioms of classical logic are epistemically reliable.
These principles have been employed in a wide variety of theoretical contexts, most notably in
mathematics, and for a wide variety of purposes and yet have remained classical. A number of non-
classical or deviant logics have been developed, some gaining a number of supporters, but what we call
classical logic has yet to be displaced. Again, I believe that our judgments regarding principles are more
reliable than our judgments about particular cases, since in the latter case our personal and pragmatic
concerns are more liable to obscure our judgments. It is from the icy heights that we make our decisive
theoretical decisions, replacing our familiar, natural, and voiced, language with an alien, formal, and mute
language.

20

Second, classical logic when extended to quantified modality produces interesting and unexpected
results that will lend direction into the investigation of Reality in itself. Interesting and unexpected results,
especially when arrived at by deductive methods on the basis of widely accepted assumptions, are
especially rare and much to be valued. Indeed, part of the reason that we can even produce an interesting
theory of Reality in itself in a roughly a priori manner consists in the fact that logic is not theoretically
neutral.
Third, I shall extend classical first-order logic to the point where it is sufficiently expressive to
express nearly all statements of the theory. Deductively speaking this permits an axiomization of
metaphysics. In this way, most of the results of the theory will be in the form of theorems produced by
axiomatic proofs (although some theorems will only admit of semantic proof). Semantically speaking,
extending logic in this way permits precise analysis, since all definitions and claims (axioms and
theorems) are interpretable in terms of a formal modal. Formal ontology (otherwise known as information
or computer ontology) presents expressive syntactically formal languages with axiomatic rules of
inference, but these languages employ too many primitive ontological categories to be provided with a
model theoretic interpretation and hence propositions about, say, predications and propositions about
events are not mutually determinative. That is, the interpretation function settling the extension of
predicates would not formally entail facts about what events occur, and vice versa.
Just as the analysis of language assists in the clarification of ordinary beliefs, given that language
use reflects commonly held beliefs, logic assists in the clarification of our common notions. We can
expect that at least some of our common notions are built into the very core of classical propositional and
quantificational logic.
12
In particular, substance ontology and classical predicate logic are mutually
reinforcing they synergize or cohere with one another. We cannot help but to make use of the subject-
predicate distinction in thought and communication. We also, cannot help but to recognize that this way
of speaking is in some respects misleading, at least upon reflection. Substance ontology might be
described as the attempt to preserve this bit of grammar, where we have no choice in the matter, along
with a robust sense of reality. In some way or other, the object, branded as a substance, is posited as the
genuine reality, while the predicates are dependent, or nothing over and above the substance. In other
words, predicates are abstracted from substance by the understanding.
While I am not a neo-positivists I concur that much of the philosophy of physics, chemistry,
biology, and perhaps even sociology is part of metaphysics, but I deny that it is the whole of metaphysics.
While I shall not assume that fundamental reality is something that will ever be directly studied by

12
Indeed, Kant obtained his categories of the understanding partly on the basis of a consideration of logical
principles.

21

physics, I shall keep track of which constraints and possibilities can be reasonably inferred from present
physics. Constraints are introduced here simply by consistency with the general scientific consensus. This
restraint may be loosened, however, when there is a well-developed and respected alternative physical
theory with which one is consistent.
Additional benefit may be gained when ones metaphysics positively coheres with natural
science. Coherence is difficult to precisely define, but what I have in mind are cases in which natural facts
might become less brute in light of metaphysics or the choice of an interpretation of a physical theory
might appear to be less arbitrary in light of metaphysics. There may, of course, be other important ways
in which metaphysics may cohere with natural science. Indeed, if ones theory demands (or strongly
suggests) a relatively alternative interpretation of physical theory but in a way that also coheres with said
interpretation, then this might be regarded as a positive (at least if said interpretation later turns out be
more adequate than its rivals).
It will be seen that careful attention to present physics can be quite fruitful even when it is not
assumed that the fundamental base is physical. For instance, in Chapter Eight observations from physics
will permit one to discern a kind of wide nomological possibility which will be identical with
metaphysical possibility just in case physicalism is roughly true. In Chapter Nine it will be observed that
careful attention to physics reveals that the observable properties of objects are not intrinsic.
7: Metaphysics and the Axiomatic Method
The axiomatic method brings together a number of the strings of the above discussion. First, if we want to
produce a theory in a principled manner, we can do no better than adopt the axiomatic method. Second,
the axiomatic method, in preventing us from relying on background beliefs when forming arguments,
forces us to adopt moderate skepticism which, I have argued, is practically necessary if we are to adopt
the view from nowhere. At each step we can only make deductions that can be traced back to the axioms.
This prevents us from taking presumed background knowledge for granted, at least when proving
theorems. Indeed, the mathematical coldness of the axiomatic method is of considerable assistance when
attempting to take up the view from nowhere.
Beyond this, the axiomatic method is the Platonic-Rationalist mode of exposition par excellence,
where Spinozas Ethics is the exemplary classic. The axiomatic method is also the most transparent and
rigorous mode of exposition. In my own judgment it is the most ideal form of theoretical exposition.
Acceptance of conclusions here need not relay on synthetic judgments where the intuitions of rival
theorists tend to differ. Synthetic judgments need only be made when motivating axioms.

22

I would like to conclude this chapter by reviewing its meta-theoretical argument.
1. The primary task of metaphysics is to produce a theory of Reality in itself (i.e., Reality as it is
independently of our conventions and practical interests).
2. It is possible for us humans, in our present state, to produce a modestly successful theory
of Reality in itself.
3. In seeking to understand Reality in itself one must strive to take up the view from
nowhere (i.e., an objective or disinterested stance toward Reality).
4. Ordinary language and beliefs about particular cases are predominately concerned with
how Reality is relative to our interests and concerns, as well as Reality as we experience it (i.e.,
with the phenomenal).
5. In taking up the view from nowhere, we must leave out considerations from ordinary
language and beliefs about particular cases.
6. Natural science does not presently provide us with a theory of fundamental physical reality
and hence does not provide us with a theory of fundamental reality simpliciter.
7. We have a concept of Reality in itself independently of physics and can develop
interesting theories about Reality in itself on the basis of common notions and logic.
8. We can and should develop theories of fundamental reality without assuming, in
advance, that fundamental reality will ultimately be the subject matter of physics..
9. We must proceed in a principled manner whereby principles are given first and various
particular conclusions about Reality are deduced, some of which may be counterintuitive.
10. The only relevant principles from which we might draw are common notions pertaining to
the nature of Reality in Itself and principles of logic.
11. The most appropriate form of exposition for a theory of Reality in Itself is the
axiomatic method.
Premise (1) is based partly on the self-descriptions of most leading metaphysicians in conjunction with
certain criticisms of the general emphasis on reductive conceptual analysis.
Premise (2) is required for the whole project to get off the ground and is, of course, almost always
presupposed in works by mainstream metaphysicians.
(3) is an inference from (1). If we want to know Reality in itself we must be maximally objective and
abstain from projecting our personal concerns and desires onto the nature of reality.
Premise (4) is made on the basis of casual observation. It follows on the assumption that fundamental
reality is bound to differ considerably from the manifest image.

23

(5) Follows from 3 and 4.
(6) is an observation on the present state of natural science.
(7) is not yet supported, but will be demonstrated throughout much of this treatise.
(8) is supported by (6) and (7). Moreover (8) is theoretically more conservative than physicalism, since
(8) simply claims that we should abstain from making a certain initial assumption.
(9) follows from (7) and (8).
(10) is inferred from (5), (6), (7) and (9) and the apparent lack of further alternatives.
(11) is inferred from (10), lack of obvious alternatives, and the presumed attractiveness of the axiomatic
method.
















Copyright Adam Labecki 2013

24

CHAPTER TWO:
SECOND-ORDER SIMPLEST QUANTIFIED MODAL LOGIC
AND POSSIBLE WORLDS

1. Sketch of SQML
+
Lexicon

Primitive Terms:
1. Object constants: a
1
, a
2
. (a, b, c)
2. Object variables: x
1
, x
2
... (x, y, z)
3. n-place predicates constants (n 0): F
n
1
,

F
n
2
. (F
n
, G
n
, H
n
)
4. n-place predicate variables (n 0): X
n
1
,

X
n
2
. (X
n
, Y
n
, Z
n
)
1-2 are object terms and 3-4 are predicate terms. In the special case where n = 0, a predicate term
denotes a proposition. For the sake of convention, however, p, with or without subscripts, will be the
variable for propositions. Object and predicates are each types and are the two most basic types.
Primitive Connectives, Operators, and Logical Connectives
1. Connectives: ,
2. Operators: !, .
3. Logical Predicates: =
4. Punctuation: (, )
Defined Connectives, Operators, and Logical Predicates
1. ^ is short for ( ) (Conjunction)
2. \ is short for (Disjunction)
3. is short for (( ) ( )) (Material Biconditional)
4. v is short for !v (Existential Quantifier)
5. is short for . (Possibility)
6. oo is short for o=o (Non-identity)
7. E!o is short for =o (Existence)
(Where o and are both object variables or are both predicate variables).
Grammar
The Greek letters , , and are used as meta-variables denoting any formula of the language , r
n
is a
meta-variable for any n-place predicate, , , and are a meta-variables for terms, and v is a meta-

25

variable for variable terms in particular. The expression (v) denotes any formula with at least one free
occurrence of v.
1. If r is an n-place predicate, where n 0, and
1

n
are object terms, then r(
1

n
) is a
formula.
2. If and are terms then = is a formula, provided that both are object terms or both are
predicate terms.
3. If and are formulas then , , and . are formulas.
4. If is a formula and v
1
... v
n
are variables, then !v
1
... v
n
is a formula.
5. Only strings that can be shown to be WFFs using 1-4 are formulas.
Logical Axioms and Rules
Propositional Logic
PL1: ( ) (Simplification)
PL2: ( ( )) (( ) ( )) (Frege's Axiom)
PL3: ( ) ( ) (Transposition)
MP: if and , then (Modus Ponens)
Uniform Substitution (US): The result of uniformly replacing any formula or formulas
1
, ,

n
in a theorem by any formula or formulas
1
, ,
n
is itself a theorem.
Classical Quantification Theory (CQT)
CQT1: !v /v (Universal Instantiation)
Provided that v and are both terms of the same type.
CQT2: !v( ) ( !v)
Provided that v is not free in .
UG: if , then !v (Universal Generalization)
Provided that v is not substituting for any term o that has been introduced by assumption.
Some Derived Quantification Rules and Theorems
UG
-
: if , then !v !v
UG
=
: if , then !v !v
PC1: !v( ) (!v !v) (!-distribution)
QI1: v !v (quantifier interchange 1)
QI2: !v v (quantifier interchange 2)
The Logic of Identity

26

I1: = (Law of Identity)
I2: = ((o) ()) (Leibniz Law)
Leibnizs Law might appear to be substantively stronger than the mere Indiscernibility of Identicals, but
this is not really the case, since an instance of might be o = , in which case it is a mere tautology
that the two are identical. The Identity of Indiscernibles is only an interesting thesis when certain
restrictions are made upon the contents of .
Axioms and Rules of S5 Modal Logic
K: .( ) (. .) (Kripkes Axiom)
T: .
S5: .
NEC: if , then . (Rule of Necessitation)
Some Derived S5 Rules and Theorems
13

K: .( ) ( )
T:
D: .
S4: . ..
B: .
DR1: if , then . .
DR2: if , then . .
DR3: if , then
DR4: if , then .
Some Modal Definitions
Df1: Contingency:
c
( ^ .)
Df2: Mere Possibility:
m
( ^ )
Modal Second Order Logic (MSOL)
MSOL1: X.!x
1
. !x
n
(Xx
1
. x
n
(x
1
. x
n
)) (Modal Second Order Schema)
Where X does not occur free in .
MMSOL1 stipulates that for any condition that might be satisfied by some ordered n-tuple of objects,
there exists an n-place predicate true of the said n-tuple, if and only if it satisfies that condition.
MSOL2: !X!(!x.(Xx x) X)
(The difference in font for second-order predication is simply for the purpose of readability).

13
The derived rules presented here are taken from Hughes and Cresswell (1996).

27

MSOL2 governs higher order predication, i.e., predication of predicates. It stipulates that if it is
necessarily the case that everything that is X is also Y, then Y can be predicated of X. Such second-order
predication is only well-behaved when the predicates in question are monadic, since must be
monadically predicated of X.
MSOL3: X
n
= Y
n
.!x(Xx Yx)
The left-to-right conditional follows immediately from I2 but the right side by itself is not sufficient to
derive the left side by I2.
Semantics
Definition of a Model
A model, M, is triple M = (W, D, I) such that:
W is a non-empty set of possible worlds, with a distinguished member w
@
, designating the
actual world.
D is a non-empty set of objects: the domain
I is a function, called the interpretation function, such that:
(a) if o is an object constant then I(o) e D
(b) if r
n
is an n-place predicate then I (r
n
) is a set of n+1 tuples (o
1
. o
n
, w) such that for
each o
i
, I(o
i
) e D and w e W.
Regarding (a), I assigns each name to a member of the domain. Regarding (b), predicate-valuations are
intensions of r
n
. I (r
n
) determines the extension of r
n
within each w such that w e W.
The variable assignment assigns each object variable to a member of D and each predicate
variable to a set of n+2 tuples (o
1
. o
n
, w, t).
The valuation-function V
M
for ML and a variable assignment , is a function that assigns 1 or 0
(true or false) to each WFF relative to each w e W according to the following conditions:
[V=] V
M
(o = ) = 1 iff [o]
M
= []
M

[Vn] V
M
(r
1

n
, w) = 1 iff ([o
1
]
M
, .... , [o
n
]
M
, w) e I (r)
[Vn[n] V
M
(r
1
2
r
1
1
, w) = 1 iff (I (r
1
1
)) , (I (r
1
2
))
[V] V
M
(, w) = 1 iff V
M
(, w) = 0
[V-] V
M
( , w) = 1 iff V
M
(, w) = 0 or V
M
(, w) = 1
[V7
x
] V
M
(!x, w) = 1 iff for every [o
1
]
M
, e D, V
M(o/x)
((o), w) = 1
[V7
F
] V
M
(!X, w) = 1 iff for every Y replacable for X, V
M(Y/X)
((Y), w) = 1
[V.] V
M
(., w) = 1 iff for every w e W, V
M
(, w) = 1
There are derived clauses specifying the truth conditions for the remaining non-primitive connectives as
one would expect.

28

Definitions of Validity
1. is valid in (W, D, I ) iff V
M
(, w) = 1 for every w e W and every assignment of .
2. is ML -valid (written
ML
) iff is valid in all ML models.
3. A set of formulas I ML-semantically-implies (written I
ML
) iff for every ML model (W,
D, I ), every w e W, and for every variable assignment , V

(,, w) = 1 for each , e I, entails

V
M
(, w) = 1.
PL, CQT, CQT plus identity, and S5 are all well known to be both sound and complete.
Moreover, the combination of these logics, known as Simplest Quantified Modal Logic (SQML) is also
known to be both sound and complete. Second-order logic is known to be incomplete relative to its
standard semantics but the above axioms can be shown to be verified by their semantic clauses semantics.
One feature included in the present semantics that is uncommon concerns the [Vn[n] clause. It
defines predicate predication in terms of the subset relation rather than the more standard type-theoretic
approach whereby the membership relation is employed. The present account is simpler and does not
require that we countenance some well-defined hierarchy of predicates. This should not be terribly
surprising since higher order logics are reducible to second order logic (Hintikka 1955).
2. The Domain of Discourse: D
In most theoretical contexts the domain of discourse, also known as the universe of discourse, is
understood to contain some specific set of objects which are relevant to the subject matter at hand. In
some mathematical contexts, D might be interpreted as the set of real numbers, in other contexts it might
be chess pieces, or spacetime points, etc. How should D be interpreted when doing metaphysics? It is
sometimes supposed that, in the contexts of metaphysics, D should be construed as broadly as possible. If
x belongs to ones ontology at all, then x ought to belong to D. This is not, however, possible. Even if
completely unrestricted quantification is coherent (and I am not certain whether it is), it cannot be the case
that everything we quantify over is a member of D. Predicates, for instance, are not members of D and if
some predicates stand for anything in ones ontology, such as universals or tropes, then one cannot
include everything in ones ontology within D. It is also uncertain whether or not we would want to
include all abstract objects within D. If we did, then D would not be a set but a proper class, since it
would include every ordinal number. But then we would not be in a position to speak of the power-set of
D, since a proper class cannot have a power-set. If so, we could not define n-predicates as members of the
power set of D
n
. Accordingly, we would be better off construing abstract objects as higher-order objects
or as having their own separate domains, such as the set of all possible worlds (to be discussed below).

29

Indeed, were we to directly quantify over possible worlds in the object-language, we would be
quantifying over members of W, not of D.
Accordingly, unless ones ontology is particularly austere, D cannot include absolutely
everything or even close to everything. So what should D be understood to include when we are doing
general metaphysics? In the very least, since we are trying to produce a theory of Reality in itself, D
should be populated precisely by those entities that are most fundamental. That is, those objects which do
not vanish when one takes up the view from nowhere. I have already suggested that those beings should
be called substances. Should D be understood to include some non-fundamental objects in addition to the
fundamental ones? That, I think, is a matter that can only be decided in the face of further theoretical
considerations.
3. Possible Worlds: W
In addition to the domain, D, and the interpretation function, I , ML includes a set W, of what we have
called possible worlds: a formula is possibly true just in case there exists some world w at which is
true (i.e., a world such that w ); a formula is necessarily true just in case it is true at every possible
world. Talk of possible worlds has generated an enormous literature in the years since Kripke (1963)
introduced possible world semantics for modal logics. This is not the place for a comprehensive survey of
that literature.
14
It is, nevertheless, necessary to provide a relatively brief account of how possible world
talk is to be interpreted.
What is a possible world? There are two general answers to this question in the literature. The
most common answer is that a possible world w is an abstract object, typically a set of propositions, such
that any world w is both maximal and consistent with respect to propositions:
Maximal: for any proposition p, w represents p or not-p.
Consistent: For any w e W, if w represents both p and q, then p and q do not contradict one
another.
There exists exactly one world, w
@
, such that for any proposition p, if w
@
represents p (i.e., if w
@
e I (p))
then p is made true by the universe, where universe is construed broadly to include the totality of
concrete being, hence w
@
is called the actual world. Since worlds are abstract worlds, the actual
world is not the universe but the one possible world that correctly represents the universe. Nevertheless,

14
Divers (2002) is the most comprehensive study regarding the status of possible worlds talk. Sider (2002) provides
a good overview of many of the salient positions. Pruss (2002) provides a somewhat different breakdown of the
different possible world stances one might take.

30

I shall often speak loosely in using actual world to denote both the one world that correctly represents
the totality of reality as well as to denote the totality of reality itself.
There are a number of ways in which one might construe abstract possible worlds. Lewis (1986a),
who calls such worlds ersatz possible worlds, considers three ways in which an ersatzer might construe
possible worlds: linguistic, pictorial, and magical.
I distinguish three principle versions [of ersatzism]: linguistic, on which ersatz worlds are like
stories or theories, constructions out of words of some language, and represent in virtue of
meanings given by stipulation; pictorial, on which they are like pictures, or scale models, and
represent by isomorphism; and magical, on which they just represent, it is simply their nature to
do so, and there is nothing to be said about how they do it. (Lewis 1986a: 141)
Lewis admits that the only such version of ersatzism that had been explicitly proposed at the time of his
writing Lewis (1986a) is of the linguistic variety, but rightly notes that those who offer no explanation of
the representational power of possible worlds might be said to default to magical ersatzism. Moreover,
these other two varieties of ersatz worlds have continued to be discussed in the literature and each has
found some defenders.
The second general approach to possible worlds is Modal Realism (MR), according to which
possible worlds are not abstract objects but concrete total ways in which things might have been. More
specifically, a world is the mereological sum of its constituents and each world is both spatiotemporally
and causally disjoint from every other world. MR is famously presented and defended by Lewis (1986a).
Lewis defense of MR consists of two major components: (a) appeal to the various ways in which
possible worlds have been put to work and (b) criticism of the various ways in which ersatz worlds might
be construed. I shall proceed to state my own favored version of linguistic ersatzism along with the
charges that Lewis has raised against it.
Since we are interested in knowing what commitments might be entailed in employing Kripkean-
style semantics, let us begin by examining them. The first thing to note is that our formal language is not
fundamentally propositional. Rather than speaking of propositions being true or false at worlds, it is more
accurate to speak of first order atomic formulas (without variables) being true or false at worlds, which
formulas are logically more fine grained than propositions. It is evident that the truth values of
compound and quantified formulas are made true by the truth-values of atomic formulas. So the worlds
need only be sets of atomic formulas (r
1

n
). That is, each world w is wholly determined by the
interpretation function I assignment of predicates to sets of n-tuples of elements of D. At any given world,
for every set of n names and any n-place predicate r, w (r
1

n
) = 1 or w (r
1

n
) = 0. Such total
predicate assignments are known as state descriptions. Lewis attributes this approach to worlds to

31

Carnap (1971/1980) and Skyrms 1981,
15
although the general idea back goes at least to Wittgensteins
Tractatus 4.26. Moreover, worlds need only include all true atomic formulas while entailing the false
ones by implication. I shall like to adopt something similar to state description ersatz worlds
State descriptions plus logical implication, however, are certainly not enough to define our
possible worlds. Or rather, they provide us only with logically possible worlds. Many of these worlds will
not even be conceptually possible, much less metaphysically possible. We do not want to have possible
worlds including round squares and married bachelors, yet these are not, strictly speaking, logically
impossible. Such things are, however, conceptually impossible. Conceptually impossible worlds can only
be eliminated if all extra-logical predicates are explicitly defined, but this cannot be humanly done and, at
any rate, goes beyond the basic idea of a state description. Hence, to be a possible world, at least in the
sense of metaphysical possibility, is not simply to be a state description. What we are after is a
distinguished subset of the set of state-descriptions.
Even if all conceptual truths were granted, however, this would not be sufficient, given the fact
that we wish to deny that conceivability entails metaphysical possibility. For instance, Lewis (1986a: 155)
raises the question as to whether or not it is possible for something to possess both positive and negative
charge. <Having positive charge> does not seem to conceptually entail <not having negative charge>.
This presents the ersatzer with two apparent problems. First, it is difficult to decide such cases in a
principled manner. Second, the ersatzer is bound to face so many cases of this kind that his theory of
modality will not be humanly axiomizable. That is to say, it is not possible for the ersatzer to fully specify
which worlds are metaphysically possible.
On MR there is a fact of the matter regarding which worlds are possible: namely all of the island
worlds that concretely exist. Hence, on MR it is clear that modal reality does not depend upon how we
construct and interpret our formal semantics.
Is there a way for the ersatzer to claim that there is a matter of fact regarding which state
descriptions describe metaphysical possibilities? The ersatzer will be in trouble here if ersatz worlds are
the ultimate truth-makers of modal claims, but the ersatzer should deny this in any case, since abstract
worlds are not the sort of things that can ground the truth about what is or is not possible. I shall latter
arrive at the position of what Contessa (2010) calls hardcore actualism (HCA) and it will be argued that
HCA performs just as well as MR in this respect. The result, however, will not be a general theory of
modality but a theory of metaphysical modality. Moreover, the range of possibilities will not be as
extensive as many might expect but I shall argue that such expectations are unjustified in the first place.

15
Skyrms Tractarian Nominalism, however, takes facts to be basic, and hence lends itself to the employment of
a propositional language rather than a full first-order language.

32

4. S5 and the Metaphysics of Modality
S5 modal logic is typically taken to be the most appropriate modal logic for use in modal metaphysics. In
order to see why, it is first necessary to say something about weaker modal systems. On modal logics
weaker than S5 one must define modal frames, where a frame is a pair (W, R) such that W is the set of
all possible worlds and R is an accessibility relation. R is a relation between possible worlds, where
w
1
Rw
2
is read w
2
is accessible from w
1
. When only the K axiom is in play, no assumptions are made
regarding which properties are satisfied by R; indeed, it is not even assumed that worlds are accessible to
themselves, which is why, on K modality, necessity at w does not even entail actuality at w. The semantic
interpretation of the T axiom is that world frames are reflexive, hence . . Likewise, the semantic
interpretation of the S5 axiom is that world frames are Euclidian.
Euclidian: (xRy ^ xRz) yRz
If a relation is both reflexive and Euclidian, then it is also an equivalence relation. Hence, by our axioms
the accessibility relation is an equivalence relation and so it can be effectively dropped. Every world is
identical with respect to what is possible and with respect to what is necessary. The result is that .
and . . are both validated by the semantics. This is essentially why S5 modality is
generally taken to be the natural logic for modal metaphysics. The intuition is that it cannot be
contingently the case that is metaphysically possible or metaphysically necessary.
One way of hashing out the idea that metaphysical necessity and possibility should be invariant
across the set of metaphysically possible worlds would be to contrast metaphysical contexts of modality
with other modal contexts. For instance, it is common to construe temporal logic upon the basis of modal
logic (Prior, 1967), where the set of all possible worlds is replaced by the set of all times. Temporally
speaking, it is not possible for the industrial revolution to begin in China. The industrial revolution
already began and began in England. This is sometimes referred to as the necessity of the past. We
say that what has been done cannot be undone. When we are speaking of metaphysical possibility, on the
other hand, we do not want tense to be a factor: it is still metaphysically possible that the industrial
revolution might have begun in China. We are trying to take up the view from nowhere. Time is, in a
sense, perspectival (at least for eternalists) just like place. Our situation in space and time makes no
difference to what is possible, nor does our epistemic state. Moreover, even if the world had turned out
otherwise than it did, its turning out differently could not have changed what is possible.



33

5. Some General Definitions
The following are a few useful definitions that can be produced on the basis of SQML
+
.
Df3: Singular Formula: is singular if and only if it contains at least one object constant.
Df4. General Formula: is general if and only if it is not singular.
Df5: Impure Predicate: For any property X, X is impure if and only if:
.!x(Xx (o
1
. o
n
))
(Where each o is an object constant)
Df6: Pure Predicate: A predicate X is pure if and only if it is not impure.
Properties may be distinguished between those which are determinate and those which are
determinable. Color, for instance, is a determinable property, while red is a determinate of color. The
distinction, however, is typically relative. For example, red is a determinable property with respect to
scarlet. W.E. Johnson (1921) provided a relatively early examination of this distinction, in which the
following four characterizations are made. First, determinate properties come in families (or classes),
where each family corresponds to exactly one determinable property. Second, determinable and
determinate properties differ in scope, since any determinable property is broader in scope (i.e., applies to
more things) than any determinate property that falls under it. Third, there are highest-level
determinables, such as color and shape which do not fall under any other determinable and which
cannot be compared to one another, since they are the ultimate dimensions of comparison. Fourth, the
determinable-determinate relation is non-conjunctive, in that one does not obtain a determinate from a
determinable by adding some third term to qualify the said determinable, in the way that equal angled
is added to triangle to yield equilateral triangle.
Searle (1959) analyzed the determinable-determinate relation along similar lines as Johnson, and
attempted to formulate a definition of determinate property. His focus is upon the non-conjunctive
nature of determinates. First, he defines more specific as a proper form of property predication (i.e.,
X is more specific than Y, if and only if, being an X implies being a Y, but not conversely).
Df7: Non-Conjunctively More Specific: A X is non-conjunctively more specific than Y if and
only if only if:
(X ^ Y) ^ !Z!W!x((Xx (Zx ^ Wx)) ^ ((Zx Yx) ^ ((Wx \ Wx) Yx)))
The smaller left conjunct ensures that X is more specific that Y, while the longer (and quite ugly) right
conjunct ensures that X is more specific in a non-conjunctive manner. More specifically, the second

34

conjunct is contrived to rule out cases in which X, Y, Z, and W have something like the following
interpretation:
Xx: x is red
Yx: x is colored
Zx: x is colored but not green
Wx: x is red or not-colored
X is equivalent to the conjunction of Z and W, but it is not the case that everything that is colored but not
green is red, nor is it the case, on this interpretation, that W or its negation entails being colored.
Searle adds a final requirement. If X is a determinable of Y, then for any Z, if Z is a non-
conjunctive specification of Y, then X and Z are logically related:
Df8: Logically Related: X and Y, X and Y are logically related if and only if:
!x(((Xx Yx) \ (Xx Yx)) \ ((Yx Xx) \ (Yx Xx)))
For instance, both red and scarlet are non-conjunctively more specific than color. Being scarlet entails
being red, so red and scarlet are logically related. Likewise, both red and green are non-conjunctively
more specific than color. Being red (all over) entails not being green (all over) and the converse is true as
well. Hence red and green are also logically related.
Finally, Searles definition of being a determinate of is as follows:
Df9: X is a determinate of Y, if and only if, (1) X is non-conjunctively more specific than Y and
(2) for any Z, if Z is non-conjunctively more specific than Y, then X and Z are logically related.
Many determinables can be analyzed relative to either some scale (which may be either a
continuous or discrete)
16
or magnitude. Even if colors cannot be identified with the color spectrum of
light, each color can nevertheless be mapped to some part of that spectrum. Likewise, tonal pitches can be
mapped to sound-wave frequencies, etc. Scientific laws deal largely with magnitudes, such as mass,
velocity, and spatial distances (along with certain constants, which are typically wholly determinate
magnitudes). In such cases, a determinable property is one that can take on various different values.
17


16
In Platos Philebus, Socrates seems to suggest that all of being can be analyzed in terms of limit and the
unlimited, where the unlimited is a kind of determinable property defined upon the basis of a continuum, and
where each point of the continuum is a determinate limit of the said determinable.
17
A fuller logic of determinable magnitudes would want to introduce some sort of logic of comparatives or degrees,
which typically assigns each determinate of a determinable to a partially order set (often the Real numbers). See:

35

Finally, wholly determinate might be defined as follows:
Df10: Wholly Determinate: For any property X, X is wholly determinate, written X, X if and
only if there is no Y such that Y is a determinate of X.




















Copyright Adam Labecki 2013

Mundy (1989). A general logic of predicates of magnitude would not concern absolute values but rather
comparative values.

36

CHAPTER THREE:
SQML AND THE THREAT OF MODAL COLLAPSE

1. The Denial of Modal Collapse (A1)
In what follows some important SQML theorems shall be presented. Many of these theorems are of great
significance. The first system of QML was presented by Barcan (1946, 1947). Her system combined
classical first-order logic with C.I. Lewis S2 and S4 modal axioms along with two additional axioms
governing the interaction between quantificational and modal operators, which formulas have come to be
known as the Barcan Formula (BF) and Converse Barcan Formula (CBF).
18
Prior (1956) both simplified
and strengthened Barcans system by combining first order logic with S5 modal logic. An important
result of this strengthening is that BF and CBF turn out to be theorems. Hence SQML, when using S5
modality, does not require any additional logical axioms aside from those already employed in classical
propositional logic, quantification theory, the logic of identity, and S5 modal logic. It is simply the
immediate result of their conjunction.
The first extra-logical axiom is to be introduced, which guarantees that the actual world is not the
only possible world.
Do1: Modal Collapse (MC): .
MC can be derived from S5 modal logic plus the following axiom:
.
19

MC is the doctrine that possibility and necessity are co-extensive.
20
It might also be called
metaphysical determinism or metaphysical fatalism. MC has had some defenders in the past, apparently

18
Williamson (2006) has noted that Movahed has observed that Ibn Sina (better known in the West as Avicenna)
anticipated both the Barcan Formula (BF) and the Converse Barcan Formula (CBF). See Movahed (2006).
19
Proof:
1. . Assumption
2. . 1. K
3. . . S5 theorem
4. . 2, 3. PL
5. . D
6. . 4, 5. PL
Henceforth, . . will be known as DR 5
20
Hughes and Cresswell (1996) have named any modal system that validates MC Triv., since the modal operators
are rendered trivial in any system in which MC is valid.

37

by Ibn Sina
21
and Spinoza.
22
MC is logically consistent; in particular, MC holds on models in which there
is exactly one possible world.
The denial of MC is typically taken for granted in modal metaphysics but, given the axiomatic
methodology of the present treatise, it must be explicitly introduced as an axiom. It cannot, of course be
stated as ( .), since that would prevent any formula from being such that it is possibly true
just in case it is necessarily true. The simplest way of expressing DMC is in terms of quantification over
proposition (i.e., 0-place predicates).
A1*: Denial of Modal-Collapse (DMC): !p(p .p)
23

A1 guarantees that there is more than one possible world. In terms of natural science, any indeterminacy
in the world will be sufficient for A1 to hold.
2. Necessity and Identity
T1: Necessity of Identity (NI): !x!y(x = y .x = y)
1. x = y (.x = x .x = y) I2
2. (x = y ^ .x = x) (.x = y) 1. PL
3. (.x = x ^ x = y) (.x = y) 2. PL
4. .x = x (x = y .x = y) 3. PL
5. x = x I1
6. .x = x 5. NEC
7. x = y .x = y 4, 6. MP
9. !x!y(x = y .x = y) 7. UG
Corollary: !X!Y(X = Y .X = Y)
Df11: Rigid Designator: A term o designates rigidly if and only if it denotes the same
individual in every possible world w in which it designates anything at all.
Semantically speaking, NI is a consequence of rigid designation. Object terms in our semantics are
interpreted rigidly in that their interpretation is invariant across possible worlds. Kripke (1972/1980)
famously argued that rigid designation entails NI:

21
Ibn Sina distinguishes between that which is necessary in itself and that which is necessitated by something else,
this is basically to distinguish that which pertains to the divine nature from everything else.
22
MC was probably implicitly held by Greek Neoplatonists such as Plotinus and Proclus.
23
I have written A1* rather than A1 because it will ultimately be replaced with a more precise formulation.

38

If a and b are rigid designators, it follows that a = b, if true, is a necessary truth. If a and b
are not rigid designators, no such conclusion follows about the statement a = b. (Kripke,
1972/1980: 3)
An important consequence of this, discussed by Kripke at length, is the inadequacy of the definite
description theory of reference, according to which the meaning of a proper name is equivalent to a
definite description of the object named, where a definite description of an object a is any description that
is true of a and only true of a.
Kripke rightly insists that descriptions play an important role in fixing reference but they do not
constitute the meaning of a name. For instance, suppose there is a gathering in which only one man is
wearing a blue shirt and that person's name happens to be Bob. Then the sentence The guy in the blue
shirt fixes the name Bob to Bob. The name Bob will not cease to refer to Bob as soon as he changes
his shirt or as soon as another blue-shirted man enters the room.
T2: Necessity of Nonidentity (NNI): !x!y(x y .x y)
1. x = y .x = y NI
2. .x y x y 1 PL
3. x y x y 2. DR5
4. x y .x y 3. DR4
5. !x!y(x y .x y) 4. UG
While the proof for NI requires only the rule of necessitation, the proof for NNI requires the DR4 rule
which, in turn, requires the B axiom. It seems counterintuitive that NI could hold while NNI does not.
NNI can fail, however in QMLs in which the symmetry between world frames fails, i.e., when the
underlying modal logic does not include the B axiom. If x and y happen to be non-identical in w
@
, there
may be some world w
j
in which x = y is true provided that it is not the case that w
j
Rw
@
, in which case
NI would be violated.
3. Existence and Necessity
3.1 The Necessity of Existence
T3: Necessity of General Existence (NGE): .x y=x
1. !y y x x x CQT1
2. x = x !x y x 1. PL
3. x = x I1
4. !x y x 2, 3. MP

39

5. x y = x 4. QI2
6. .x y = x 5. NEC
NGE is simply the result of applying the rule of necessitation to the CQT theorem x y=x. Since
E!x is short for y y=x, NGE tells us that, necessarily, something or other exists. Hence, it could
not have been the case that there is nothing at all. It is well known that CQT is not compatible with the
existence of empty domains, which has perhaps been the primary motivation for the adoption of Free
Logic (to be briefly discussed below).
T4: The Necessity of Existence (NE): !x.y(y = x)
1. .x y=x NGE
2. !x.y(y = x) 6. UG
First Corollary: !X.Y(X = Y)
Second Corollary (NNE): .!x.y(y = x)
Third Corollary: For any cardinal k, if k-many objects exist, then, necessarily k-many objects
exist.
The second corollary here is of especial importance, since it tells us that anything that even possibly
exists necessarily exists.
In terms of possible world semantics, NE entails that every x that exists in the actual world is
such that, in every possible world, there exists some y with which x is identical. Apparently, then,
everything necessarily exists. This is, of course, an incredibly strong theorem.
NE is not, however, as strong as it might seem at first blush. For instance, the fact that there are
Kangaroos does not entail that there are necessarily Kangaroos. That is, while .xKx can be inferred
from the conjunction of NE and Ka we cannot thereby infer x.Kx. That is, just because there is
necessarily something that happens to be a kangaroo, it does not follow that anything is necessarily a
kangaroo. Hence {!x.y(y = x), Ka, xKx} is a consistent set. The inference to x.Kx,
and hence also .x.Kx, requires an additional assumption to the effect that !x(Kx .Kx). That is
to say, one is inclined to make certain essentialist assumptions when evaluating the implications of NE
and, given such assumptions, NE certainly does seem unacceptable. So the conflict is between classical
logic and essentialism. The methodology of this investigation dictates we side with logic.
NE denies a thesis that Nelson (2009) has named Absent:
Do2: Absent: xy x = y
T5: Denial of Absent: xy x = y

40

1. !x.y x = y NE
2. x.y x = y 1. Def.
3. xy x = y 2. Def.
Absent is the doctrine that there exists something such that it might not have existed. The denial of
Absent is, of course, highly counter intuitive. While it is hardly more controversial to suppose that
abstract objects necessarily exist than to suppose that they exist at all, it seems obvious that just about any
concrete object might not have existed. For instance, one might not have been born if ones parents had
not met and procreated. Broadly speaking, it seems contingent that the Earth exists at all, much less that it
turned out to be a life-supporting planet. Indeed, it seems to be contingent that the universe should even
support life or, for that matter, physical aggregates. But, again, all of these concerns arise from certain
background assumptions that go beyond what is being said in the logic.
T6: Transworld identity: There exists some object x and worlds w
1
and w
2
such that x exists
in w
1
and x exists in w
2
.
Proof: Transworld identity is an immediate consequence of A1, NGE, and NE. By A1 there are
at least two possible worlds, w
1
and w
2
, by NGE some x or other must exist, and by NE, we know
that x exists in every possible world. In particular, we know that x exists in both w
1
and w
2
.
According to transworld identity, when we ask whether Socrates might have been a carpenter and not a
philosopher, we are asking if there are any possible worlds that represent Socrates himself as a carpenter
but not as a philosopher. That is to say, Socrates is represented as existing at every possible world in
which he may be considered (and by NE he can be considered in every possible world).
24
Transworld
Identity is generally taken to be the alternative of Lewis Counterpart Theory, according to which all
(concrete) objects are world-bound, but have transworld counterparts. The counterpart of Socrates at
some world w is not Socrates himself but some concrete existent resembling Socrates in some pertinent
manner. When we ask could Socrates have been a carpenter? we are likely asking if someone very
much like Socrates, living in very similar circumstances, might have become a carpenter. Hence,
counterpart theory takes a distinctly qualitative approach to apparently de res modal claims.
At first blush counterpart theory seems quite sensible. If someone, in the actual world, very
similar to Socrates became a carpenter, it would be reasonable to infer that Socrates might have become a
carpenter. But here is the problem as I see it. It is not because someone very similar to Socrates did
become a carpenter that makes it the case that Socrates could have. It is because of the way Socrates is (at

24
It should be noted, however, that we have not really established that even possibly Socrates exists. But some sort
of examples must be given when discussing these theorems. To anticipate a doctrine according to which it is not the
case that Socrates exists, it might be that Mereological Nihilism is true and that nothing like a soul or immaterial
mind exits.

41

least at some time during his life) that makes it the case that he might have been a carpenter. He just
happens to share this way with some other person who did, indeed, become a carpenter. I see no reason
why this complaint should vanish when the appropriate similarity relation holds between Socrates and
someone in another possible world rather than someone in the actual world. Hence the counterparts of
MR seem to be more like epistemic justifiers of modal claims than metaphysical truth-makers but with the
peculiarity that these epistemic justifiers are epistemically inaccessible to us and hence do nothing to
justify our modal claims.
The precise meaning of transworld identity claims cannot be spelled out until we have discussed
the relation between powers and modality, but to put it schematically, I will say that there is a world in
which Socrates became a carpenter just in case Socrates possesses the potency to become a carpenter.
That is, des res modality about an object x would be grounded in the dispositional properties of x
Modal realism comes into conflict with SQML, in that SQML commits us to transworld identity
while, as mentioned above, Lewis opts for counterpart theory.
25
Lewis (1986a) argues that combining MR
with transworld identity leads into trouble with the Identity of Indiscernibles (I2). McDaniel (2004),
however, has suggested how MR might be combined with transworld identity provided that one embraces
certain non-Lewisian doctrines such as endurentism and mereological pluralism. Accordingly, transworld
identity does not quite block the way to MR, but it blocks the conjunction of modal realism and four-
dimensionalism.
Do3: Necessitism is the doctrine according to which NNE is a logical truth.
Do4: Contingentism: Contingentism is the doctrine that NNE is not a logical truth.
26

Evidently, endorsing SQML commits one to affirming Necessitism and denying Contingentism.
It would seem, then, that possible worlds differ only with respect to their predicate extensions.
One might have drawn this conclusion simply be looking at the semantics of , since nothing is said
about worlds being assigned different subsets of D. In quantified modal logics with varying domains,
each possible world w is assigned its own domain: dom(w), and the relevant semantics clauses are
modified accordingly.
27
Just as S5 modality makes world frames unnecessary, SQML makes it
unnecessary to assign a domain of individuals to each possible world. Hence we arrive at the surprising
conclusion that the number of existing objects is a strictly necessary matter. Since worlds (i.e., state
descriptions) do not differ in terms of what exists, they differ only in terms of which objects fall under the
extension of which predicates. To cite Williamson:

25
David Lewis (1968).
26
These two terms are coined by Williamson (2010).
27
In particular: [V7
x
] becomes: V
M
(!x, w) = 1 iff for every o e dom(w), V
(o/x)
(, w) =1.

42

What objects there are and their identity, distinctness, and number are all necessary matters.
Indeed, in LPC=S5 any proposition expressible in terms solely of identity, negation, conjunction,
the universal quantifier and necessity is either necessarily true or necessarily false. Contingency
has its place within a necessary framework of objects (Williamson 1998: 269).
This permits us to provide an alternative formulation of A1 (although the best version still awaits):
A1**: !X
n
!x
1
!x
n
(X
n
x
1
x
n
.X
n
x
1
x
n
)
It will be noted that, on NNE, predicates exist in every possible world. In particular, any predicate
X
n
will exist even in worlds in which the extension of X
n
is empty. Given A1, there will presumably be
some predicates which are not actually instantiated.
Also, inasmuch as propositions may be construed as 0-place predicates, every proposition p
necessarily exist. This does not mean that every proposition is necessarily true (which would immediately
result in MC), since p p is a perfectly grammatical formula. Every proposition necessarily exists as
something we can refer to and talk about, as in the case of ordinary predicates.
3.2 Truth-Makers and the Truth-Maker Principle
Df12: Truth-Maker: For any formula and objects x, x is the truth-maker of just in case the
existence of x makes it the case that is true.
28

Do5: Truth-Maker Maximalism is the doctrine that, for every true formula , there exists some
x such that is true in virtue of the existence of x.
Do6: Truth-Maker Necessitism is the doctrine that, for any formula and any object x, if, in the
actual world, x is sufficient to make true, then in every possible world in which x exists, is
true.
Do7: Truth-Maker Principle (TMP): .( x.(E!x ))
In words: necessarily, for any true formula , there exists some x such that, necessarily, if x exists then
is true. It is natural to suppose that true formulas must be made true by something and, in the literature,
such things have come to be known as truth-makers.
29
The doctrine that truths require truth-makers
might be regarded as a contemporary form of the correspondence theory of truth, according to which any
true formula is true in virtue of corresponding to the world in some appropriate way. Truth-maker
theory is a specific kind of correspondence theory of truth; in particular, the truth-maker theorists claims

28
In articulating Truth Maker one might prefer to use plural rather than singular variables.
29
The best known truth-maker theorist has been Armstrong (1997, 2004). Other prominent truth-maker theorists
include Martin (1996) and Cameron (2005, 2008b). The contemporary discussion of truthmakers was initiated by
Mulligan, Simons, and Smith (1984).

43

that whenever a proposition p is true, there exists some thing (or things), in virtue of which p is true. That
is to say, for the truth-maker theorists, truth does not merely supervene upon how the world is (or the
conjunction of what exists and how those things are) but upon which things exist. Truth-makers are
typically thought of as facts or (Armstrong-style) states-of-affairs.
30

At first glance, Tarskian-style semantics might seem to imply some sort of correspondence theory
of truth, but truth in a model is typically regarded as a set-theoretical surrogate for full blooded truth
makers.
31
That is, the model-semantic conception of truth is neutral regarding the question of truth-
makers.
It is natural for the truth-maker theorist to be a truth-maker maximalist. If one believes that some
true propositions require truth-makers, it is natural to suppose that all truths require truth-makers. One
difficulty that arises for the truth-maker maximalist, however, is that it is not clear what the truth-makers
for negative propositions could be. What makes it true that there is no dragon under this table? Is it
merely the absence of a dragon under this table?
32
But then we are treating such an absence as if it were a
thing or a positive feature of the world. In particular, Lewis (1992) has suggested that we not seek truth-
makers for negative existentials. Nevertheless, there have been various attempts to preserve Truth-Maker
Maximalism in the face of such challenges. Maximalism is also threatened by some necessary truths, such
as mathematical truths. What makes it the case that, for every set S there exists some set S such that the
cardinality of S is greater than the cardinality of S? One could opt for some sort of trivial truth-makers
here, but this seems to betray the spirit of truth-making theory, since truth-making is not supposed to be a
trivial matter.
While truth-maker theorists are somewhat divided regarding maximalism, the great majority
support Truth-Making Necessitism.
33
The idea is that, if x is sufficient to make true, then, in any world
in which x exists, is true.
34

TMP, then, is a principle governing the combination of truth-maker maximalism and truth-maker
necessitism. As Williamson (1999) has observed, it is not compatible the conjunction of NE and A1.


30
This is Armstrongs (1997, 2004) position, but one could seek other sorts of truthmakers, such as tropes.
31
See: Tarski (1944) and Mulligan, Simons, and Smith (1984).
32
Armstrongs favored solution is to appeal to a maximal state of affairs, which would effectively contain every
other state of affair as a part. So the proposition there is no dragon under the table would be made true by the fact
that the maximal fact does not include the state of affairs there is a dragon under the table.
33
Although Heil (2000) is a truth-maker theorists who denies necessitism.
34
Armstrong (1997) presents an argument in favor of Truth-Maker Necessitism but Cameron (2008b), who happens
to be a proponent of the doctrine, argues persuasively that Armstrongs argument, and any evident modification
thereof, ultimately begs the question.

44

T7: The Truth-Maker Principle Entails MC
35

1. .( x.(E!x ) TMP
2. x.(E!x ) 1. T
3. x(.E!x .) 2. K
4. ( ^ x.E!x) . 3. PL
5. x.E!x NE
36

6. x.E!x 5. PL
7. ( ^ x.E!x) 6. PL
8. . 4. 7 PL
TMP gets into trouble here precisely because truth becomes a matter simply of what objects exists. In
particular, when states-of-affairs (i.e., property instantiations) are denoted by terms, it turns out that
nothing could have differed from the way things actually are.
The denial of TMP does not mean that we must reject every truth-maker theory as such.
Moreover, we may still follow Lewis (1992) in proclaiming that truth supervenes upon being, where
being is construed broadly so as to include not only what there is, but also the ways things are. The way
that some object happens to be is not, itself, some further object.
It is possible to quantify over facts or states of affairs provided that they are not treated as terms
but as formulas. Again, the problem with Armstrongs approach to Truth-Makers and states of affairs is
that he treats them as thing like in a way that encourages quantifying over them as terms. As
Williamson (1999) argues, the difficulty might be elevated if we permit ourselves to quantify over
propositions, formulas. Again, p p is a perfectly grammatical formula. Accordingly, Williamson
(1999) suggests the propositional route by replacing TMP with TMP*, which also happens to be a
theorem of SQML
+
.
T8: TMP*: .( p(p ^ .(p ))
1. .( ) PL, NEC
2. ( ^ .( )) 1, PL
3. ( ( ^ .( ))) p(p ^ .(p ) 2. EI
4. p(p ^ .(p )) 2-3. PL
5. p(p ^ .(p )) 4, CQT, PL
6. .( p(p ^ .(p ))) 5, NEC

35
The proof is run using singular variables, but could also be run using plural variables.
36
Recall that E! is a logic predicate defined as y x=y.

45

The problem with TMP*, however, is that we are effectively using propositions as truthmakers for other
propositions, which is especially evident in the following variation of TMP*:
T9: TMPU: .!q(q p(p ^ .(p q))
Propositions (or sentences) are truth-bearers, they are the things that need to be made true. In the very
least TMP* will not permit a reductive analysis of truth whereby truth might be said to supervene upon
being or to be grounded in non-truth-bearing facts. A modified version of TMPU might provide a
principle of grounding, whereby some propositions obtain in virtue of others.
I shall have occasion to return to the matter of grounding later; as for truth-making, expressing a
reductive theory of truth-making in the object-language may be possible but it would be cumbersome. It
would be necessary to have a theory of non-linguistic facts or events which make propositions true. But
such facts or events are going to have to have conditions under which they obtain anyway, and so there
will need to be truth-makers for formulas about their obtaining. The object language does not allow us to
say something along the lines of snow is white just in case snow is white. Moreover, NNE cannot be
said to be to blame here, since E!x is itself a truth (or falsity) bearing formula.
3.3 Direct Reference
Df13: Definite Description: is a definite description of x, written :x, just in case
!y((y/x) y = x)
Do8: Definite Description theory of Meaning: The definite description theory of meaning is the
doctrine that the meaning of any object x is precisely a definite description of x.
Do9: Direct Reference: The doctrine that the meaning of any object term o is precisely the
element of D to which o is assigned.
A matter that is closely related to, and easily confused with, rigid designation is that of direct reference.
Given the combination NI, NNE, and A1, the definite description theory of reference is likely to fail. It is
however, epistemically possible for there to be a definite description of each individual that holds rigidly
across possible worlds and this description might even be compatible with a tremendous degree of modal
flexibility. This would be the case if each object possessed a unique set of essential properties. This is
plausibly the case for abstract objects but it seems unlikely to be true of concrete objects. Moreover, even
if it is true of concrete objects, it will later be seen that, such descriptions will almost certainly be elusive,
and hence, it is not clear that such descriptions could serve as the meanings of names. This would be a
problem for descriptivists, since most are internalists regarding the meaning of names.
One might try to argue in favor of direct reference by appeal to semantics. The meaning of an
object constant is determined by the interpretation function I while the meaning of an object variable is

46

determined by the variable assignment . Both constants and variables are assigned directly to elements
of D. Hence the direct theory of reference is practically built into the semantics. The question is whether
the content of a theory of reference is exhausted by its model semantics. To be sure, one can produce
models in which names are not be rigid and hence do not directly refer. The standard Tarskian treatment
of names, which is part of the semantics for SQML, however, seems to favor direct reference.
3.4 NNE and the denial of Alien
Perhaps more important than NE is NNE, which is a corollary of NE obtained simply by applying the rule
of necessitation. In terms of possible world semantics, NNE entails that, in every possible world w,
everything necessarily exist; that is to say, everything that possibly exists necessarily exists. Accordingly,
SQML is only compatible with constant domain semantics every possible world contains the same
objects. Not only do we tend to suppose that actual things might have failed to exist, we also tend to
presume there could have been individuals that do not in fact exist. Hence NNE appears to deny what
Nelson (2009) has named Aliens:
Do10: Aliens: xy(y = x)
The here is being employed as the actuality operator. Aliens is the doctrine that there might have
been something that does not actually exist. Semantically speaking, it is the doctrine that in some possible
world, there exists something that does not exist in the actual world. As for the actuality operator, Zalta
(1999) presents the logic of actuality in terms of the following two axioms:
LA1:
LA2: .
Formulas involving the actuality operator are logically true but contingent and hence the rule of
necessitation cannot be employed on formulas containing such an operator. The valuation for is:
[V] V
M
(, w) = 1 iff V

(, w
@
) = 1
T11: Denial of Aliens: xy x=y
1. y x=y CQT. LA1
2. .y x=y 1. LA2
3. !x.y x=y 2. UG
4. .!xy x=y 3. BF
5.xy x=y 4. Def. and
Despite the fact that NNE is a theorem of SQML, it would be rash to suppose that something has
somehow gone wrong with SQML such that it should be construed to suggest MC. First, the S5 modality

47

underlying SQML clearly does not validate MC and the interaction between S5 and CQT validate neither
of the following formulas:
x .x
x x.
The metaphysical question, then, is: how can NNE be true and MC false? The holding of MC is
sometimes called necessitarianism (Nelson 2009). So another way of putting the question is how can
necessitism be true while necessitarianism is false? How can it be the case that everything that could
possibly exist necessarily exists while it is nevertheless the case that the world might have been different
in, presumably, countless many ways? It seems that any metaphysics that affirms NNE while denying MC
will be strongly revisionary in one way or another.
4. Contingentism and the Problem of Possibilia
While the denial of Alien may be counterintuitive, its affirmation also has counterintuitive implications,
since it is would seem to commit one to Possibilia:
Df14: Possibilia: x is a possibilia, written x, if any only if:
y(x = y ^ x = y)
T10: The Class of Possibilia is Empty: xy(x=y ^ x=y)
1. !x!y(x=y x=y) LA1, PL, UG
2. .!x!y(x=y ^ x=y) 1. PL, NEC
3. !x.!y(x=y ^ x=y) 2. CBF
4. x.y(x=y ^ x=y) 3. Def.
Semantically speaking, possibilia is doctrine that there exists some x such that [x] e D but it is not the
case that [x] e (d)w
@
.
While possibilia have not been universally shunned by modal metaphysicians, they are shunned
more often than not. Many philosophers, beginning with Parmenides, have denied the intelligibility of
referring to that which does not exist. What exactly is wrong with possibilia? First of all, when we speak
of possibilia, we seem to be saying that there are things that do not exist but might have existed, or, in
other words, that there are non-actual things. The required distinction between bare existence and actual
existence, however, seems to be dubious. In the very least, if there are things that are non-actual (and if
MR is false) then such thing would seem to be rather shadowy creatures we are not quite sure just
what it would mean for something to be a possibilia.

48

While Aliens does not directly entail that there are possibilia, it is difficult to understand how it
could express the idea that alien objects exist according to other worlds without forcing one to quantify
over merely possible objects. Divers (2002: 211) presents the following question for actualists about
possible worlds:
D-problem: how can actualists semantically construe formulas such as: xy(y = x)?
Stalnaker has recently produced a monograph dedicated to this problem:
How, on an actualist interpretation of possible worlds as ways a world might be, is one to account
for the possibility that there be individuals other than those that actually exist? That is the main
focus of this book (Stalnaker 2012: ix).
Divers notes three ways in which the actualist may try to circumvent the D-problem. One may (i) adopt a
QML with constant domains, namely SQML, (ii) construe non-actual possibile individuals as
descriptions, or (iii) posit a set of individual essences that exist in every possible world but which are only
instantiated in some.
Perhaps the most promising way to be both an actualist and a contingentist is to construe
contingently non-existent objects as abstract objects, namely as definite descriptions. A name denoting a
merely possible object will be construed not so much as referring to an object that does not really exist,
but as referring to some description. There are various ways in which one might construe such items. I
shall not go into precise details regarding how such objects can be accommodated, since such
accommodations involve a number of semantical complications. It is worth noting, however, that there
are two general approaches one could take. One could construe non-actual objects as structured abstract
objects belonging to D or as properly non-referential but meaningful in that each is effectively assigned
some set of properties and relations.
It is generally thought that such ersatz individuals fall prey to at least two problems. First, as
noted in Lewis (1986a) it would deny the possibility of there being indiscernible (actually non-existent)
objects. If the meaning of a term denoting an ersatz individual consists in some kind of description and if
two names are assigned the same description, then they must be thought of as referring to the same
possible individual. Second, as noted by McMichael (1983), ersatz individuals raise a problem when
interpreting formulas with iterated modalities, such as the following:
x(Fx ^ Px)
Suppose F is interpreted as fathered, while r is the name Ratzinger and P is interpreted as plumber.
Hence the formula can be read to say in some world w
1
there is some x that is fathered by Ratzinger and
which, in some world w
2
, is a plumber. Given that the meaning of x is some description and supposing

49

that x is not a plumber in the first mentioned world, it seems that we fall into a contradiction, since the
description will include both Px and Px.
One way to respond to both of these difficulties would be to assign each non-designating name a
set of duples {F, w,F, w, }, where the first term of each member is a one place predicate and the
second is a possible world. Hence, a for non-denoting term a, Fa would be true at a world w just in
case F, w e I (a). The problem of indiscernibility at a particular world will be overcome, since two
names will refer to the same ersatz individual just in case they refer to the same transworld description.
While it may be possible for two distinct individuals to be contingently indiscernible, it is implausible to
suppose that two distinct individuals could be necessarily indiscernible. Likewise, transworld descriptions
permit transworld identity for ersatz individuals.
The limitations of this proposal, however, become apparent when we turn to the evaluation of n-
place predicates, where n >, 1 involving at least one non-actual object and at least one actual object. We
cannot evaluate such formulas in the usual Tarskian manner, nor can we evaluate them in the manner of
the previous paragraph. A general approach to evaluation is required that holds for both real and ersatz
objects while still permitting one to precisely distinguish the two classes.
Antonelli (2000) provides non-Tarskian proto semantics for positive Free Logic with a single,
actual, domain. The proposed semantics satisfy compactness and completeness relative to its
axiomatization, although it is rather complicated and unorthodox. Roughly speaking, non-actual objects
do not denote but they are effectively assigned predicates by a proto-interpretation function. The
semantics are capable of evaluating formulas containing both real and ersatz objects while still
distinguishing them, but the semantics are intended for non-modal free logic and hence could not be
directly employed in the modal case without further modification. I presume that, with sufficient
ingenuity, the required modifications could be made but suspect that the resulting model theory would not
only be unorthodox and complicated but also oblique. That is, I suspect that any formally adequate
solution would do little to provide elucidate of the meaning of formulas containing alien objects.
The third general response to the D-problem involves the introduction of haecceities. Plantinga
(1974), for instance, introduces a set of necessarily existing individual essences. Individual essences may
be defined as follows:
Df15: Individual Essences: A property E is an individual essence just in case
x(Ex .((y x = y Ex) ^ !y(Ey y = x)))

50

A property E is an individual essence just in case (i) there is a world w and an object x such that E is
exemplified by x at w, and at any world w, if x exists at w then x exemplifies E at w and (ii) at any world
w and any object y, if y exemplifies E at w, then y = x.
Each individual has exactly one essence and each essence is the essence of exactly one individual.
Individual essences are necessary existents and, in the formal semantics provided by Jager (1982), D is
precisely populated by such essences. In this respect, introducing haecceities (or essences) is very similar
to necessitism, only one has populated the domain with the essences of individuals rather than with the
individuals themselves. Even if Socrates did not exist, the individual essence, Socrateity, or the
property of being Socrates would still exist. Socrates is a contingent existent precisely because his
individual essence is exemplified in some worlds but not in others.
Linsky and Zalta (1994) complain that this revision to semantics is costly, since the objects
themselves are no longer directly expressed in the semantics. First-order quantifiers are interpreted as
quantifying over subsets of D (whichever subset is assigned to the world of evaluation) but such items are
not individuals per se but properties. Accordingly, a notion of co-exemplification must be introduced in
order to discuss what is usually understood as exemplification simpliciter. For instance, Socrates is snub-
nosed must be interpreted as Being-Socrates is co-exemplified with being-snub-nosed. As Menzel
(2008) notes, many of Plantingas definitions cannot be analyzed in terms of the formal semantics that his
theory requires.
There are, moreover, some philosophical concerns one might have with the individual essences of
non-actual objects. As properties, individual essences will not be qualitative, since in that case they would
be no different than descriptions. If they are non-qualitative, it seems they must be construed along the
lines of being identical with Socrates or simply being Socrates (Menzel 2008). In the first case, it
seems quite clear that the property of being identical with Socrates is meaningless unless Socrates
actually exists. Indeed, both ways involve the singular term Socrates, the meaning of which depends upon
reference to Socrates. More generally, Plantingas theory of individual essences appears to entail the
denial of existentialism:
Do11: Existentialism: .!p(p(a) x a=x)
Existentialism is the doctrine that for any singular proposition p including a singular term a, if p exists
then a must also exist. On a Russellian theory of propositions, the object named by a is literally a
component of p, and hence any proponent of Russellian propositions will be an existentialist. One,

51

however, need not be a Russellian in order to endorse existentialism. Anyone endorsing the direct
reference theory of names (or something near enough) will also endorse existentialism.
37
At any rate, it
does not seem that the property being identical with Socrates or being Socrates can exist unless
Socrates himself exists. Indeed, one might have noticed that the very definition of haecceity leads one into
troubles, since in the first clause we still find ourselves quantifying mere possibilia: a property E is an
individual essence only if there is some x such that, at some world, x exemplifies E. It should also be
noted that, on Jagers formalization, BF, CBF, and NE are blocked only at the cost of banishing
individual constants and hence a significant loss of the expressive power of QML. Given necessitism, of
course, existentialism is necessarily true:
T11: Affirmation of Existentialism: .!p(p(a) x a=x)
1. x a=x CQT
2. p(a) x a=x 1. PL
3. .!p(p(a) x a=x) 2. UG, NEC
It is especially difficult for a linguistic ersatzer be a contingentist while also being an
existentialist. Given existentialism and contingentism, singular propositions will exist contingently. If
possible worlds are sets of propositions (or atomic formulas) then possible worlds, even as abstract
objects, will be contingent. Suppose that world w, contains some proposition p that does not exist at a
world w, we should, I think, conclude that w is not accessible from w and that it is inaccessible precisely
because it does not exist at w. If this is right, no world containing alien objects will be accessible and
hence possible relative to the actual world, which simply means that it is not possible for something to
exist if that something is other than anything that actually exists. That is to say, one will be limited to
modal semantics in which D may contract but not expand. The necessity of existence will fail, but it will
not be the case that anything actually non-exist object might have existed.
But perhaps I am moving too fast here. There may be ways in which actualists worlds may be
contingent without wreaking havoc on our modal semantics. Stalnaker (2012) has proposed an algebraic-
based semantics for contingent propositions which is designed to confront these difficulties. His proposal
involves a tactical retreat from both extensionality and realism regarding possible worlds. Stalnaker
follows McMichael (1983) in his understanding of extensionality and realism in this context.
Extensionality is lost due to the fact that not all instances of modal operators will quantify over the same
set of possible worlds W (note that we require variable domain of worlds rather than a variable domain of

37
Marcus (1986) argues along these lines.

52

individuals). It will be non-realist in that the semantics will contain elements of D which are supposed to
represent possibile individuals while denying that any such element is something like a mere possibilia.
Note that the failure of full-on realism here is not with respect to concrete possible worlds but with
respect to ersatz worlds. Stalnaker concedes that these features of his theory may be unattractive but he
regards the theoretical costs to be worth it if we can avoid the ontological costs of possibilism or the
metaphysical revisionary costs of necessitism. Indeed, decisions must be made and each choice seems to
come with a price.
It will be admitted that Stalnakers preference for theoretical costs (i.e., addition of complexity
and loss of elegance) over ontological and metaphysical costs is quite sensible. After all, why should we
suppose that the right theory will be relatively simple and elegant? Perhaps we have an aesthetic bias in
favor of such theories (in fact I am sure that we do) but why should we expect the world to be such that it
permits such theories to be the best theories? The history of science lends some credence to our
preference for simplicity and elegance but good science can also get very complicated and messy,
especially the biological sciences.
At any rate, it is evident that, given necessitism, strict actualism can be easily accommodated
without any theoretical sacrifices.
T12: Strict Actualism: !x(y x=y y x=y)
1. y x=y .y x=y NE, PL
2. .y x=y y x=y T
3. y x=y .y x=y 1, 2 PL
4. y x=y y x=y LA1
5. y x=y y x=y 3, T
6. y x=y y x=y 4-5 PL
7. !x(y x=y y x=y) 6 UG
Semantically speaking, Strict Actualism is the doctrine that for every x, if [x] e D, then [x] e
(dom)w
@
.
38
Necessitists can also easily satisfy the following doctrine:
Do12: Serious Actualism: .!xX(Xx y x=y)

38
This definition of actualism is somewhat stronger than is the norm but I think that Williamson is correct in his
judgment that the actualism vs. possibilism, as typically formulated is rather obscure (Williamson 1998, 2010).
Nevertheless, the above definition of actualism follows Adams (1981) where he writes Actualism is the doctrine
that there are no things that do not exist in the actual world (7). Barcan (1986) and Bennett (2006) also endorse this
position.

53

Serious actualism is the doctrine that objects can only have properties in world in which they exist.
T13: Serious Actualism: !xX(Xx y x=y)
1. y x=y CQT
2. Xx y x=y 1. PL
3. !xX(Xx y x=y) 2. UG
4. .!xX(Xx y x=y) 3. NEC
In the absence of NNE, it is awkward to deny serious actualism and it is also awkward to reject it. Even if
one has adopted a deviant logic under which the proof for strict actualism fails, it is alarming to say that
objects can have properties at worlds in which they do not exist. If the contingentist, however, adopts
strict actualism, then his QML will have truth-gaps and worlds cannot be maximal under entailment.
The necessitist, however, is spared from this dilemma.
5. The Barcan Formulas
T14: Barcan Formula (BF):
39
!x. .!x
1. !x. . CQT1
2. !x. . 1. DR3
3. . S5 Theorem
4. !x. 2, 3. PL
5. !x. !x 4. CQT2
6. !x. .!x 5. DR4
First Corollary: x x
40

Second Corollary: !X. .!X
41

Third Corollary: X X
Apparently BF and its converse, CBF, were initially formulated and endorsed by Ibn Sina, better
known in the West as Avicenna. Hence it might be appropriate to call it the Ibn Sina-Barcan Formula.
42


39
The following proof was first put forward by Prior (1956).
40
Proof:
1. .!x !x. BF
2. !x. .!x 1. US [/]
3. .!x !x. 2. PL
4. !x !x 3. Def.
5. x x 4. QI2
41
Williamson (2009) shows that this second-order version of BF holds even on quantified modal logics with
variable domains.
42
Williamson (2006) cites Movahed (2004), in which this discovery is presented.

54

Nevertheless, I shall continue to employ BF as its abbreviation. BF can be read as follows: if, in the
actual world, every entity necessarily satisfies some condition , then it is necessarily the case that all
entities in every possible world satisfy . The problem here is that it seems that there should be possible
worlds, perhaps even this one, in which everything is essentially (and hence necessarily) a physical being,
but from that it would follow that everything in every possible world is a physical being.
Of course, given NE, this result should not be at all surprising. If Aliens is false and if everything
that exists is necessarily a physical being, then it is impossible that there should be any non-physical
beings. In Chapter Eight I shall defend a conception of metaphysical modality relative to which this result
is not at all surprising: if there is nothing that has the potential to become non-physical or to generate
something that is non-physical then it is just not possible for anything to be non-physical.
The existential form of BF, x x, states that, if in some possible world, there exists
something that is , then there is something in the actual world that could have been . For example,
suppose that it is possible for something to be a goblin. On BF

it follows that there is something in the
actual world that in some possible world is a goblin. This means that BF conflicts with sortal essentialism
(SE). Recall that essentialist assumptions also run into trouble when interpreting NE and NNE. By NE, if
kangaroos are essentially kangaroos, then every possible world includes all of the kangaroos present in
the actual world. Moreover, by NNE every world has precisely those kangaroos present in the actual
world. Accordingly, T. Parsons (1995) has written the following regarding BF:
Antiessentialism is required as follows. Certainly there might have been more porcupines than
there are. So there must be possible worlds in which there are things that are not porcupines in this
world ... but the additional things will not be porcupines in the actual world. We thus need the
possibility of a thing that is not a porcupine in this world [but] is one in the other world; that thing
cannot essentially be or not be a porcupine (Parsons 1995: 11).
43

Likewise, Williamson (2010):
What there is should not be understood as what kinds are instantiated. The necessitist may
agree with the contingentist that it is contingent whether there are rivers. The dispute concerns
whether it is contingent or necessary which particular things there are, never mind what kinds they
instantiate (Williamson 2010: 8).
Assuming that there really are such things as rivers, when the necessitist says the St. Laurence
necessarily exists, it is the existence of the particular object, which we call the St. Laurence, that is
necessary, not its being a river.
Sortal essentialism is difficult to define in a non-circular manner. At present I will make no
attempt to provide a rigorous definition of substantial sortal. All that I shall say is that sortals in general

43
It might be noted that Parsons himself reads NE to merely say necessarily, everything is something rather than
to say that everything necessarily exists. This, however, requires the rejection of Existentialism.

55

are understood to be something like count nouns as opposed to mass nouns and sortal essentialists
typically have in mind those sortals which denote substantial kinds.
Regarding count nouns and mass nouns, you can ask someone how many bricks are needed but
you cannot ask how many clays are needed; indeed, clays is not even an English word. So, in general,
we might say that sortals are count-kinds rather than mass-kinds.
Regarding substantial kinds, some might only admit the basic kinds of physical fields or particles.
Most sortal essentialists, however, will include the various kinds of living creatures and some might
include things like planets, stars, and artifacts as well. Using as a primitive predicate for substantial
sortal, sortal essentialism may be stated as follows:
Do: Sortal Essentialism (SE): !X(X !x(Xx .Xx)
44

Given NNE, SE would entail that every substantial sortal that is instantiated is necessarily instantiated
and, indeed, necessarily instantiated by just those things that actually instantiate it. Moreover, it would be
impossible for there to be alien sortals, sortals that are not instantiated in the actual world but
instantiated in other possible worlds. While the conjunction of NNE and SE does not immediately lead to
MC, it certainly places severe restrictions on what might have been the case. The world could not have
differed enough to change the general course of biological evolution or which sorts of artefacts were
invented!
45
Hence, it seems, contrary to appearance, genetic mutation would not be genuinely random,
nor the evolution of the cosmos.
Moreover, it will later be demonstrated that classical logic, when extended to time entails
permanentism, being the doctrine that everything that exists always exists. Given permanentism, SE
entails that the universe never changes with respect to which ultimate sortals are instantiated. This leaves
us with the following disjunction: either none the sortals that we observe with our senses or which are
studied by the natural sciences are substantial sortals or SE is false.
BF also appears to conflict with origins essentialism, according to which everything necessarily
has the origins that it has in fact.

For instance, despite the fact that Wittgenstein had no son, presumably
Wittgenstein could have had a son. According to BF, it follows that there exists something in this world

44
Any doctrine or definition that relies upon non-official primitives shall not be numbered, since they are not
officially part of the formal theory. Their use here is for purely dialectical purposes.
45
Some essentialists might deny that sortal essentialism applies to artifacts by denying that artifacts are genuine
substances.

56

that, in some possible world, is Wittgensteins son.
46
Taking origin as primitive, origins essentialism
may expressed as follows:
Do: Origins Essentialism (OE): !x!y(Oxy .Oxy)
47

Origins essentialism is the doctrine that, for every x and every y, if x originates from y, then x necessarily
originates from y. Were OE the case, not only would every possible instance of a sortal be a necessary
instance, but every possible circumstance in which a sortal comes to be instantiated would be a necessary
circumstance. OE, however, might be vacuously satisfied by permanentism, since it would not seem that
necessary objects would even have origins at all (at least not temporal origins, which is what OE
advocates surely have in mind).
T15: Converse Barcan Formula (CBF) .!x !x.
1. !x CQT1
2. .(!x ) 1. NEC
3. .(!x ) (.!x .) K
4. .!x . 2-3. MP
5. !x(.!x .) 4. UG
6. (!x(.!x .)) (.!x !x.) CQT 2
` 7. .!x !x. 5-6. MP
48

First Corollary: x x
Second Corollary: .!X !X.
Third Corollary: X X
CFB is valid even upon the weakest classical QML, combining first order logic and the modality
of system K. Prima facie, CBF is uncontroversial. In its universal form it says that if, in every possible
world w, is true of everything in the domain of w, then is necessarily true of everything in the actual
world. This seems to be obvious truth. The denial of CBF is equivalent to: .!x ^ x. On
possible world semantics, this says that, in every possible world, everything is such that it satisfies and

46
It is rather surprising that that Quine (1953.1961: 156) took SQML in to entail Aristotelian essentialism (all that
he in fact shows is that it involves a theorem equivalent to the necessity of identity. Parsons (1969), however, rightly
shows that SQML does not entail any kind of interesting essentialism.
47
One might prefer to use a plural variable in the place of y.
48
Employing CBF enables the following alternative proof for NE (Deutsch 1990):
1. !xy x=y CQT
2. .!xy x=y 1. NEC
3. .!xy x=y !x.y x=y CBF
4. !x.y x=y 2, 3. MP
5. .!x.y x=y 4. NEC)

57

yet, in the actual world, there exists something that is possibly not . Thus, there is something such that,
in some possible world is not and hence the second conjunct contradicts the first.
The existential form of CBF, x x, states that if is possibly true of something in
the actual world then, in some possible world, there is something that is . Again, this seems to be
obviously true by the very definition of possibility. Moreover, just as the denial of its universal form
entails an absurdity, the denial of the existential form of CBF entails: x ^ x: there exists
something that might have satisfied and yet it is not the case, in any possible world, that there exists
something that satisfies . As Williamson (1998) rightly notes, it is difficult to conceive how CBF could
possibly fail.
Semantically, however, CBF entails that there could not have been fewer individuals than there in
fact are. Fewer, if course, is a matter of cardinality, and so CBF does not directly entail NE.
Nevertheless, BF and CBF jointly entail NNE.
49

Bird (2007) reads BF to suggest that possibility is fully grounded upon actual things that exist.
Taking its existential form, BF tells us that if it is possible for something to satisfy a given condition, then
there is something that could have done so. He suggests that the x in the consequent possesses some
dispositional property or properties such that, under certain counterfactual circumstances, would result in
x satisfying the condition in question.
Given, CBF, however, we know that the material conditional goes both ways. If something
possibly satisfies then, it is possible for something to satisfy . Simchen (forthcoming) argues that, for
those of us who share Russells robust sense of reality, it is more natural to ground general (de dicto)
modality upon particular (de res) modality. Furthermore, he argues that the best reason to accept BF
arises not from considerations of modal logic but from how it provides focus on our analysis of
metaphysical modality:
If something is generally possible, then it is only so because of a corresponding possibility for
something in particular. As an illustration, let us consider the question of what makes it the case
that it is possible that there be a chair occupying an actually empty corner. The default intuitive
answer is that what makes this generally possible is that something might have been a chair
occupying the corner presumably one of the actual chairs might have been such. And for that to

49
Williamson (2010) provides the following axiomatic proof that the conjunction of BF and CBF entail NNE:
1. xy x=y xy x=y (CBF)
2. xy x=y (CQT)
3. xy x=y (2. NEC, Def. )
4. xy x=y (1,3. PL)
5. xy x=y (4, S5)
6. x y x=y x y x=y (BF)
7. xy x=y (5, 6. PL)
8. .!x.y x=y (7. Def. )

58

be the case, presumably a counterfactual history ending with a chair in the corner would have had
to diverge at some point from the actual history of the empty corner, a counterfactual history
involving one of the actual chairs. (Simchen forthcoming).
While I am not quite convinced that this use of BF provides the best reason for accepting it, I do,
however, agree that it is preferable to ground general modality in particular modality and that BF does
suggest as much. The fact that it is possible for something or other to satisfy does not seem to explain
why some particular might possibly satisfy but the fact that some particular might satisfy does seem
to explain why it is that something in general might satisfy .
Later I shall introduce a principle, A3, that effectively entails that modality must be grounded in
the actual, provided that objects possess powers (and eternalism is true). Simchens reference to possible
histories is also of significance for what is to come, since possible worlds will ultimately be construed as
possible, alternative, histories of the actual world.
6: Some Additional Theorems of SQML
+
T16: x. .x
T17: !x !x
T18: X
n
.X
n
1. X .!x
1
x
n
(X
n
x
1
x
n
x
1
x
n
) MSOL2
2. X .!x
1
x
n
(X
n
x
1
x
n
x
1
x
n
) 1. PL
3. .!x
1
x
n
(X
n
x
1
x
n
x
1
x
n
) ..!x
1
x
n
(X
n
x
1
x
n
x
1
x
n
) S4
4. .!x
1
x
n
(X
n
x
1
x
n
x
1
x
n
) X 1. PL
5. ..!x
1
x
n
(X
n
x
1
x
n
x
1
x
n
) .X 4. K
6. .!x
1
x
n
(X
n
x
1
x
n
x
1
x
n
) ..!x
1
x
n
(X
n
x
1
x
n
x
1
x
n
) S4
7. .!x
1
x
n
(X
n
x
1
x
n
x
1
x
n
) .X 5-6. PL
8. X
n
.X
n
2, 7. PL
9. .X
n
X
n
T
10. X
n
.X
n
8-9 PL
7. Alternatives to SQML
In this work I have made the conservative axiomatic decision to adopt SQML and have begun to
explore some of the metaphysical implications of this decision. Nevertheless, a few things might be said
regarding alternative forms of QML which do not validate NE, BF, or CBF. In what follows, the two

59

primary alternatives to SQML shall be discussed: Kripkes Variable Domain Quantified Modal Logic
(VDQML) and Free Modal Logic (FML).
Until recently, the most common alternative to SQML has been to adopt variable domain
semantics, first introduced by Kripke (1963). Variable domains permit one to represent the idea that
different possible worlds can include different individuals. Semantically speaking, this involves only a
minor complication, as one only needs a function dom that assigns a subset of D to each world w. While
the semantics of VDQML remains elegant enough, its axiomatization is both complicated and weakened.
CQT1 (Universal Instantiation) is jettisoned and use of the rule of necessitation (NEC) on open formulas
is prohibited. CQT1 is not valid because, given a world w and a variable assignment , might not
belong to the domain of w.
50
The standard proofs for each of the controversial SQML theorems involve
the use of CQT1 or applying NEC to open formulas. These same changes, however, also make the proof
theory more cumbersome.
In VDQML CQT1 is replaced with the following generality theorem, whose standard form is as
follows: !x(!y x/y) provided that x is substitutable for y in . Nevertheless, NE, BF, and CBF
can still be derived if one is permitted replace free occurrences of variables with constants. Thus Kripke
must banish the use of constants altogether. Regarding this, Menzel has written: It is surely a sad irony
that a system whose motivation is to capture our modal intuitions most notably, intuitions about
contingency cannot so much as permit us to talk about specific contingent individuals and say of those
individuals that they are contingent (Menzel 2008). In this way, we avoid saying that individual
constants are necessary only because we cannot say anything about individual constants.
51
In short,
VDQML turns out to be more difficult and less expressive than SQML and, while it enjoyed favor for
many years, it has lost much of its luster. The decline of VDQML has partly been to the benefit of SQML
but it has also been to the benefit of what is currently SQMLs main rival: Free Modal Logic.
Arguably the most promising alternative to SQML involves adopting Free Logic (FL), in which
CQT1 does not obtain because the domain of objects is understood to be possibly empty. There are three
different approaches to construing the semantics of Free Logic: positive, neutral, and negative. On the
positive semantics, atomic formulas containing empty terms can be true. For instance, a = a is true
even if a does not exist (hence the I1 axiom is retained without restriction, but existentialism and strict
actualism fail). In this case D is typically said to have an inner-domain of objects that are said to exist,

50
Williamson (1998) 262.
51
It is also ironic that the great advocate of the necessity of identity should banish the use of names from QML. We
can express the idea that if x any y are identical, then they are necessarily identical, then they are necessarily
identical, but we immediately go beyond the confines of logic when we say Hesperus is identical with
Phosphorus.

60

and an outer domain of non-existing objects. On neutral FL semantics, atomic formulas containing empty
terms are not assigned any truth-value. This approach becomes complicated, however, when attempting to
evaluate complex formulas containing both empty and nonempty terms and there is disagreement
regarding which such formulas should be given truth-values and, if so, how these values should be
assigned. On negative FL semantics, atomic formulas containing empty terms are regarded as false.
In Free Modal Logic (FML) positive semantics are adopted, where an outer domain is interpreted
to include objects that do not exist in the actual world, but may exist in other possible worlds. This
permits one to speak of non-existent things, such as fictional or otherwise non-actual entities without
thereby quantifying over them. That is, we can say that Santa Clause is a jolly old man without saying
that there exists some x such that x is a jolly old man. Axiomatically speaking, FML accomplishes this by
replacing CQT1 with !x (y y=a a/x), which reads: if everything is , and a exists, then a is .
Unlike VDQML, FML permits the use of constants and does not restrict the use of necessitation. Various
forms of FML have been developed and defended by, among others, Kit Fine (1978) and Garson (2001).
So why not adopt FML rather than SQML? In addition to the fact that the proof theory for FML
is more difficult than that of SQML, other difficulties crop up. For instance, substitution instances of co-
extensive open formulas typically fail on any semantics for free logics with identity (Nolt 2010).
Additionally, FML with an outer domain requires that we quantify over possibilia and hence is not
compatible with strict actualism, serious actualism, or existentialism. FML without an outer domain may
be possible in principle but would technically difficult. These criticisms are by no means intended to
present a decisive case against FML, but rather to highlight some features of FML that might be regarded
as unattractive.








Copyright Adam Labecki 2013

61

CHAPTER FOUR:
ON PREVIOUS INTREPTETATIONS OF SQML

In the present chapter the formal axiomatic presentation is paused so as to consider and criticize two
previous interpretations of SQML.
52
I shall make use of some terms which will not be formally defined
until later in this work. Most such terms, however, are fairly familiar and, I will generally provide some
explanation of their meaning.
The first interpretation to be discussed is due to Linsky and Zalta (1994, 1996) and Williamson
(1998, 1999, 2010).
53
This interpretation attempts to preserve certain essentialist intuitions and intuitions
regarding the contingency of existence by revising the categories of the concrete and the abstract. In
particular, it is claimed that being concrete is a contingent matter: any concrete object could have been
non-concrete and there are many non-concrete objects that could have been concrete. Accordingly, I shall
refer to this position as CCNC, short for the contingency of concreteness and non-concreteness.
The second interpretation is due to Nelson (2007, 2009) and is based upon an anti-essentialist
interpretation of QML proposed by Stalnaker (1975). This interpretation rejects essentialism outright
while appealing to our intuitions regarding the contingency of existence by understanding all concrete
objects as bare particulars, all of which are modally identical to one another. As such, I shall refer to
this interpretation of SQML as BPAE, short for bare particular anti-essentialism.
1. Exposition of Contingency of Concreteness and Non-concreteness
Let us begin with some core definitions that are put forward by CCNC theorists:
Concrete: For any x, x is concrete, written Cx if and only if x exists in spacetime.
Abstract: For any x, x is abstract, written Ax, if and only if .Cx.
Even those who deny sortal essentialism will typically believe that concreteness is an essential
predicate,
54
that is, it is generally thought that, for any x, if x is concrete then x is concrete in every
possible world in which x exists. Moreover, the intension of abstract is typically taken to be the
complement of the intension of concrete, that is, abstract and non-concrete are usually taken to be
synonymous. According to CCNC both of these platitudes are mistaken. It is denied that abstract is the

52
Dialectically speaking, it would be been preferable for this to be the fourth chapter, but the evaluation of said
theories involves drawing upon some definitions that could not be introduced by that point.
53
Parsons (1995) defends a similar interpretation of SQML.
54
See, for instance, P. Mackie (2006).

62

complement of concrete, so being non-concrete does not entail being abstract, and neither concrete
nor non-concrete are essential predicates.
Where the contingentist says that Aristotle might not have existed, the CCNC necessitist says that
Aristotle might not have been concrete. Likewise, when the contingentist says that there might have been
things that do not actually exist, the necessitist says that there are non-concrete things that might have
been concrete. While the number of objects that exist could not have differed, the number of concrete
objects might have differed. In this way CCNC explains our intuitions regarding the contingency of
existence.
CCNC may be characterized by way of the following definitions:
Possible-X: For every x and every X, x is a possible-X, if and only if Xx is true.
Merely Possible-X: For every x and every X, x is a merely-possible-X, if and only if Xx ^
Xx is true.
Bare Possibilium: For every object x and every world w, x is a bare possibilium in w just in case
x is merely possibly concrete in w.
The Contingency of Concreteness and Non-concreteness (CCNC): CCNC is the doctrine that
there are contingently concrete objects and bare possibilia (i.e., contingently non-concrete
objects).
Despite the name, bare possibilia are not really possibilia. Possibilia are merely possible objects;
objects that could have existed but happen to be non-existent. Let t be a merely-possible table, t is not a
table that happens to be non-actual, rather, it is an actual thing (it actually exists) that is not a table but it
could have been a table. There are no non-concrete tables, but there are non-concrete things that might
have been tables had they been concrete.
According to the CCNC necessitist, the contingentist has implicitly restricted his quantifiers to
concrete existence (or perhaps concrete and abstract existence, while excluding bare possibilia), under
which restriction there would apparently be a plethora of counterexamples to NE and BF.
While CCNC denies ordinary essentialism, Zalta (2006) proposes an alternative version of
essentialism, according to which a property X is essential to an object x if and only if x is X in every
world in which x is concrete.
55
Let us call this sort of essentialism weak essentialism.

55
He actually uses an existence predicate, E!, but interprets it to mean concrete existence.

63

Weak Essential Property: For any property X, X is a weakly essential property if and only if:
.!x(Cx Xx)
The primary metaphysical axiom introduced by CCNC is the doctrine that concreteness is a
contingent predicate. One result of this ideology is that the universe is populated by, presumably, a vast
number of bare possibilia. A question that will naturally arise in the mind of someone who has been
introduced to this doctrine is: just what are contingently non-concrete objects like? Linksy and Zalta
respond to this question as follows:
So, in answer to the above question, what are the objects required by the BF like in this world, we
respond with the usual intuitions philosophers have concerning nonconcrete objects: the properties
they have at this world are the same ones that numbers, sets and other abstract, nonconcrete
objects have. . We just appeal to the same intuitions actualists are prepared to use when
describing ordinary (essentially) abstract objects. However, we should note that contingently
nonconcrete objects have different modal properties than essentially abstract objects (Linsky and
Zalta 1994: 446).
Hence, while contingently nonconcrete objects are not, strictly speaking, abstract, their difference from
abstract objects is wholly modal, that is, they differ only in terms of how they might have been had they
been concrete.
Williamsons account of what bare possibilia are like is essentially the same as Zalta and
Linskys. Using a merely possible mountain as an example, he writes:
If one tries to describe it without reference to the merely possible, one will find little to say: it is
self-identical, it lacks spatial location, it is not a mountain. What distinguishes a merely
possible
attributive
56
mountain m1 from a distinct merely possible mountain m2? If one is forbidden to
refer to the merely possible (to use modal notions), perhaps one can only say that they are distinct
(Williamson 2000: 204).

There is nothing about the way that a merely possibly concrete object is that distinguishes it from an
abstract object. Nor are there any non-modal qualitative differences between any two bare possibilia
which would ground the modal differences between them. They differ only in terms of their modal
properties or capacities. Indeed, Williamson speaks of such modal properties as potencies:
On the view defended here, an object is essentially a locus of potential. How far it actualizes its
potential may be a radically contingent matter. But the existence of that object with that potential
is wholly noncontingent (Williamson 2002: 250-51).
Not only is it the case the existence, identity, and distinctness are necessary matters, the potentials
possessed by objects are also necessary matters. Indeed, if such potentials were not necessary matters it
would be difficult to see how the S5 axiom could hold. Contingency is wholly a matter of the degree to

56
By possible
attribuitive
Williamson means possible as it is read in a formula like x:Fx : there exists
something that might have been (and perhaps is) F. This is in contrast with possible
predicative
, which possible as it
is read in a formula like :xFx : something possibly exists and is F.

64

which potentials are realized. Presumably the idea here is that in different worlds, different circumstances
obtain, resulting in potentials being realized to different degrees. The contingency of realization is
radical in that an objects potency need not be realized at all.
If contingently non-concrete objects differ from abstract objects in that the former possess non-
realized potencies for concreteness while the latter do not, then the difference between the two is
dispositional. This is clear in the case of an object a that, in w
@
is a bare possibilium at time t in that it is
not yet concrete at time t but is concrete at some later time t. Compare a with some b, where b is a bare
possibilium at every time in w. It is clear that there are dispositional difference between a and b at t
because we know that a will later become concrete in response to some circumstance c while b will not.
If, however, this dispositional difference is not grounded upon any qualitative or categorical
difference, then contingently non-concrete objects possess bare dispositions, the manifestations of which
consist in their becoming concrete.
2. Critical Evaluation of CCNC
Various objections have been raised against CCNC. I shall discuss several objections that are found in the
literature while adding some of my own. First, weak essentialism does not seem to be a genuine form of
essentialism and is otherwise an arbitrary restraint. Second, bare possibilia seem to be proxies for mere
possibilia. Third, CCNC introduces an objectionable form of primitive modality. Fourth, on CCNC it is
not possible that the whole of reality should depend upon the whole of concrete reality. Fifth, CCNC
seems to be contrived so as to have as little impact as possible on our pre-theoretical understanding of
reality.
The adequacy of weak essentialism has been challenged by the likes of Hayaki (2006) and I am
inclined to agree for the following reasons.
57
Essentialism is typically motivated by the intuition that for
any actual object x, x could only have been so different while being that very thing. Essentialists will insist
that it is illegitimate (if not meaningless) to stipulate a world w in which the Sun is a jelly-fish. For natural
kind essentialists, if F is the infima species of a, then F constitutes what a is. This claim is perhaps
fleshed out in D. Wiggins (1980/2001) famous claim that identity statements are meaningless unless they
are tied to some sort of sortal. On this view, a = b is an elliptical statement that needs to be
supplemented with a sortal term, the full statement being a is the same F as b (which might be written
as a =
F
b. According to Wiggins, the very identity of an object is tied to its instantiation of some

57
Hayaki (2006), however, argues along largely different lines. She notes that Weak Essentialism one cannot even
articulate the notion that something is necessarily concrete, since .(Cx Cx) is trivial.

65

natural kind sortal. Since existence is defined in terms of identity, natural kind essentialists believe that
the very existence of an object is tied to its infima species.
Likewise, L.A. Paul describes essential properties as modal persistence conditions (L.A. Paul
2006: 334), by analogy with temporal persistence conditions. Just as one might set limits on how much an
object can change over time without ceasing to exist, a term can only rigidly designate an object over so
much modal variation.
Origins essentialism also seems to link the essential predicate in question with the very identity of
a thing. Kripke, for instance, rhetorically asks: How could a person originating from different parents,
from a totally different sperm and egg, be this very woman? (Kripke 1972/1980: 113). On this view, the
origins of a contingent object are necessary individuating conditions for it.
One might regard weak essentialism as a translation schema between CCNC propositions and
genuine essentialist claims, but it is certainly more than that, since CCNC could be true without any
interesting instances of weak essentialism holding. Indeed, on closer examination, weak essentialism
appears to be rather suspicious. Suppose that anything which is actually a horse is a horse in every world
in which it is concrete and let h be one such horse. Of course, there were times in the past in which there
were no horses, so h was non-concrete at those times, but was a merely possibly horse. When h was non-
concrete it had none of its characteristic horse properties, since each of those properties can only be had
by something concrete. So when h was not yet a horse, its actual state was no more similar to that of a
horse than that of a quasar, yet it has the possibility of being a horse while failing to have the possibility
of being a quasar.
The case of origins essentialism is even more artificial since, on its intended interpretation, it will
entail that nothing ever gets to be concrete more than once. Let h be some actual human. By CCNC in
conjunction with origins essentialism there is a unique sperm egg pair s, e such that h became concrete
in virtue of s fertilizing e and it is impossible for h to be rendered concrete in any other way. Not only
does this mean that h must be made concrete in the same way in every possible world in which it is
concrete, it also means that h can have no more than one stint of concreteness in any given world.
58
What
reason have we to believe this to be the case? To be sure, we would not be able to track Socrates
existence before and after the life we associate with the name (and in virtue of which that name has fixed
its reference), but that alone does not justify the inference that Socrates could not have been concrete prior
to or after that life. Moreover, given that most putative sortals only exist for a short span of cosmological

58
I say this is so on the intended interpretation, since even if both forms of essentialism hold, it would still be
possible that someone is born more than once to the same s/e pair, provided that both s and e get to instantiate
sperm and egg more than once and s happens to fertilize e more than once. Nevertheless, I doubt that the CCNC
weak essentialist intends to include such possibilities.

66

time, it seems arbitrary to presume that a necessary existent like Socrates could only have been concrete
while humans inhabit this little rock we call Earth.
Or rather, the restriction appears to be arbitrary unless one has decided in advance that
necessitism should involve as little revision as possible to the domain of the concrete. This appears to be a
policy of Williamson (2010) and I shall return to this matter in a moment.
The actualists qualifications of CCNC have been questioned by Bennett (2006), who argues that
bare possibilia
59
are mere proxies for possibilia. Nelson and Zalta (2009) respond to this criticism and at
least succeed in showing that CCNC adheres to the letter of actualism, since bare possibilia exist and are
actual, although its adherence to the spirit of actualism remains questionable. Despite the fact that bare
possibilia actually exist, Bennett (2006) and Hayaki (2006) are right to suggest that bare possibilia are
introduced merely as the bearers of modal properties and hence resemble mere possibilia re-described as
actual yet non-concrete objects.
A notable feature of CCNC, quite apparent in Williamsons writings, is the need for a particularly
strong form of primitive modality, the endorsement of which is sometime called modalism. Nearly all
actualists will appeal to primitive modality in some way or other. For instance, Contessa (2010) endorses
what he calls hardcore actualism, according to which modal claims are to be grounded in the dispositions,
laws, or essences within the actual world. This sort of primitive modality is arguably innocent in that it is
reasonable to believe in such modal features on grounds of scientific realism and I shall develop a theory
of hardcore actualism in Chapter Eight. Moreover, even the modality involved in dispositions should not
be entirely primitive. We believe that concrete objects can do things, but we generally believe that an
object can do what it does in virtue of the way that it is. Armstrong (1968) argued against the behaviorism
of Ryle (1949) on grounds that the latter posited behavioral dispositions that were not grounded in any
actual categorical state of the mind, which Armstrong takes to be identical with the brain.
Bare possibilia suffer from the same problem as brute behavior. They are said to possess
potentials for concrete manifestations but these potentials are not grounded in their actual states of
being. As Bennett (2006) notes, contingently non-concrete objects have no actual properties accounting
for the differences in their modal properties. This is very much like saying that two objects are perfect
duplicates with respect to their natural categorical properties, yet one has the disposition to dissolve in
water while the other does not.
I cannot believe that any object without any actual natural properties can possess the potency to
manifest natural properties under some stimulus condition. Moreover, it is at least natural to judge that

59
Although she does not use the term bare possibilia since she is specifically addressing Linksky and Zalta, who
do not use the term.

67

anything possessing a natural property must be a concrete object and so I cannot believe that there is any
such thing as a bare possibilium. The guiding intuition here is that any state S of object x in virtue of
which x has the potential be in some concrete-entailing state S, must itself be a concrete state of some
kind. In short, CCNC provides us with no reason to believe that it is metaphysically possible for an object
can pass from non-concrete existence to concrete existence.
In Williamson (2010) it is admitted that CCNC is incompatible with the following doctrine:
Chunky: Chunky is the doctrine that everything depends upon the concrete.
Typically, dependence is defined as follows:
Dependence: depends upon if and only if: .( ) is true.
The reason CCNC cannot satisfy Chunky is that many possibilia will not depend upon some particular
concrete state of affairs. Williamson seems to have moderately remote possibilities in mind here. In some
cases we can point to the grounds of a certain bare possibilium. For instance the possibility of a certain
mammal might depend upon the existence of a certain egg-sperm pair. Likewise, bare possibilia a and b
might depend upon the existence of two knives, whose handles and blades might have been exchanged.
To be sure, we can track some possible bare possibilia concretions in this way, but we should not even
say that these possibilia depend upon such states of affairs. They are necessary existents after-all, and so
there are bound to be possible worlds in which a possible mammal m cannot be traced to any state of
affairs in those worlds, since there are many possible worlds in which there are no sperms or eggs.
Accordingly, the existence of bare possibilia should be independent of concrete states of affairs
altogether. If anything the concrete turns out to be dependent on the non-concrete, since everything, or
nearly everything, will be non-concrete before it becomes concrete. Indeed, this suggests that the non-
concrete is more fundamental than the concrete. But how could objects which differ from abstract objects
solely in terms of their modal (or temporally future) properties provide the very grounds of concrete
reality? In my judgment, any objects which are to serve this function need to be more substantive than
bare possibilia (i.e., they must be full blown substances).
CCNC is motivated by the desire to defend necessitism while making minimal revisions to our
ordinary understanding of the domain of the concrete. Prima facie this is sound methodology. After all,
when defending what is already a counterintuitive philosophical position one wants to present it in the
most appealing light possible. In other words: do not multiply controversies beyond necessity! Moreover,
a defense of necessitism as such should not get mixed up the defense with distinct doctrines, even if they
happen to cohere well.

68

The problem, however, is that necessitism is a game changer. If one is really willing to accept or
entertain necessitism one should be willing to accept or entertain a revisionary metaphysics of the
concrete world. That is, to seriously entertain necessitism is to seriously entertain the epistemic possibility
that our ordinary common sense and quasi-scientific (i.e., toy model physics) conception of the world
provides us with little grip on fundamental reality. The fact that CCNC has found few converts, despite its
continued endorsement by so prominent a figure as Williamson, is evidence of this: the intersection of
those who are theoretically conservative and those who are disposed to adopt necessitism is very small
indeed. In any case, if one is going to entertain necessitism it would be unnatural not to entertain how it
would work should it be the case that concreteness is a necessary property.
My main complaint here is that CCNC theorists proceed according to general principles when it
comes to logic but when it comes to matters of substance appeal to our beliefs about particular cases. The
informal (and certainly non-deductive) argument of Chapter One arrived at the conclusion that we
should proceed in principled manner in both respects, based on the presumed unreliability of our ordinary
beliefs when applied to Reality in itself. CCNC permits one to avoid revisions to ordinary beliefs at the
expense of making numerous revisions in metaphysical principles: concreteness is contingent, the non-
concrete can become concrete, the non-concrete does not depend upon the concrete and may even ground
the concrete, there are absolutely primitive modal properties and brute modal difference between non-
concrete objects, etc. To be sure, one can consistently present a necessitist theory that involves minimal
revision to the domain of the concrete, but the result comes off as stilted and artificial.
3. Exposition of Bare Particular Anti-Essentialism
It was noted above that, weak essentialism aside, CCNC, in denying that concreteness is a necessary
property, is incompatible with just about any form of traditional essentialism. Bare particular anti-
essentialism (BPAE), presented in Nelson (2007, 2009) is a position that fully embraces the anti-
essentialist implications SQML. BPAE was first presented by Stalnaker (1979/2003), not as a variety of
necessitism but as a strategy for making QML in general safe for Quinean anti-essentialists.


This theory holds, roughly, that for every individual and every property, there are possible worlds
in which the individual has that property and possible worlds in which it does not (Stalnaker
1979/2003: 72).
60

An object satisfying the above condition can be rightly called a bare particular precisely because its
existence is independent of any property it might have, or just about any property, since some
qualifications must be added. Stalnaker adds three qualifications, which are taken from Marcus (1967)

60
The page number is from the papers appearance in Stalnaker (2003).

69

and Parsons (1969). The first qualification is that it is permissible for x to essentially possess X provided
that it is true that every x essentially possesses X. For instance everything is essentially self identical,
everything is either X or not-X, etc. The second qualification is that referential properties may be
essential, in particular if a and b are names such that a = b is true, then .a = b. In other words, NI
is not objectionable. The third qualification is that essential world-indexed properties are essential; that is,
being snubbed-nosed-in-the-actual-world is an essential property of Socrates.
61

While not an advocate of necessitism, Nelson (2007, 2009) has suggested that BPAE is a better fit
with necessitism than CCNC. He rejects the claim that concreteness and non-concreteness could be
contingent, but beyond this, BPAE is a strong form of anti-essentialism. With respect to the three
qualifications noted above, Nelson only regards the first as essential, calling his anti-essentialism
indiscriminate essentialism. It is indiscriminate in that it only admits essentialists claims that do not
discriminate one object from another. Let us call this theory BAPE-IE. BPAE-IE can be expressed in the
form of a single axiom (where quantification is understood to be restricted to concrete objects):
BPAE-IE: x. !x.
62

BPAE-IE says that if anything is such that it necessarily satisfies some condition then everything is
such that it satisfies . BPAE-IE is equivalent to the following (contrapositive):
BPAE-IE: x !x
In words: if something is possibly then everything is possibly . Just as S5 modality renders all worlds
equivalent with respect to possibility and necessity, BPAE-IE renders all objects equivalent with respect
to possibility and necessity.
As Nelson (2009) notes, intuitions regarding the contingency of existence can be accommodated
by BPAE-IE. When we say that it is possible that Aristotle might not have existed, we usually presume
that Aristotle is necessarily human. On BPAE-IE Aristotle would have existed even if there were no
humans, in which case he would have been something, but certainly not human. Likewise, if it is true that
there might have been a unicorn, it is not because there is some bare possibilium that might have been a
unicorn but because some concrete object, that is actually something other than a unicorn, might have
been a unicorn. Indeed, anything might have been a unicorn. I might have been a unicorn. Likewise,
anything might have been a son of Wittgenstein, even Wittgenstein himself, since NI is rejected.

61
Such properties, when indexed to the actual world, are formally handled by the actuality operator.
62
Nelson (2009) calls this the Reduction Axiom.

70

While Nelson rejects NI, I believe that BPAE is more attractive if it can be rendered compatible
with SQML. Indeed, it is not clear how one could block NI without also blocking NE. Again, the central
case in favor of necessitism is that it is a consequence of SQML and so we should want to preserve NI.
Starting with BPAE-IE, I suggest replacing the meta-variable with the second-order predicate
, which is intended to denote is a natural property. Natural property will be formally defined and
discussed in Chapter Six. For now it will suffice to say that they are the sort of properties that cut nature
at the joints.
BPAE: !x!X(X (x.Xx !x.Xx))
Since being identical with x is not a natural property, BPAE does not produce any conflicts with NI.
Anticipating the deduction of permanentism (which will be presented in the next chapter), much
of the modal flexibility of bare particulars is likely to be accompanied with comparable temporal
flexibility. Let permanentism be the doctrine that everything every object that exists for any time in a
world w must exist for every time t in w. M. Sullivan (2012a, 2012b) advocates permanentism
independently of necessitism.
63
While Sullivans permanentism is effectively CCNC restricted to the
temporal case, CCNC combined with BPAE suggests that objects will fall under different sortals or
natural kinds at different times. Indeed, if it is impossible for an object to be concrete without
instantiating some natural kind, observation will lead us to conclude that, in the actual world, concrete
objects instantiate different sortals at different times. Nelson characterizes sortal instances of objects as
modes:
The individuals we normally speak of people, rocks, chairs, and automobiles are really just
modes of underlying substances who can survive without those individual-modes (Nelson 2007:
466).
Such modes are similar to what Wiggins (1967) calls phase sortals. Phase sortals are sortals that
something has only for some phase, or period, of its existence, such as kitten, or policeman. Phase
sortals are contrasted with substance sortals, which are thought to be possessed for the entire duration of

63
Sullivan (2012a) discusses how classical tense logic entails permanentism. In Sullivan (2012b) permanentism is
provided with additional philosophical motivation. Her primary argument is based on the following two
assumptions: (1) if objects do come to be and perish, then it would be often be vague whether or not something has
come to exist or has ceased to exist and (2) it is impossible for existence to be vague. (1) would, of course, only
hold for composite objects, so one could resist permanentism by adopting mereological nihilism (and perhaps, also
by unrestricted composition).

71

an objects existence, such as felis catus and homo sapiens. On BPAE those sortals which are typically
regarded as substance sortals will turn out phase sortals.
64

Traditional Aristotelian substance ontology emphasizes the whatness of things the essence of
a thing consists in the kind of thing that it is. In particular, the substantial sortal, or natural kind, of an
object is definitive of what a thing is.
BPAE stands in stark contrast with sortal essentialism and its Aristotelian tradition in that it does
not link infima species predicates directly to the existence of individuals. The traditional essence/accident
distinction is rejected. Just as being an infant, being red, and having a mass of 5 kg are regarded as modes
or ways in which one and the same thing might be or fail to be, being a human, being an oak tree, and
being a gas giant are regarded as modes or ways in which one and the same thing might be or fail to be. I
shall latter pick up on this idea, construing natural phase sortals in terms of a certain kind of process that
an object undergoes.
4. Critical Evaluation of BPAE
Given that BPAE combines necessitism with bare particularism, otherwise known as the
substrate theory, standard objections to bare particularism will also hold for BPAE. A standard
objection to bare particulars is that they are objects without any properties and that this is impossible.
65

Most defenders of bare particularism, however, do not claim that bare particulars are altogether without
properties. On BPAE, bare particulars are only denied necessary natural properties. In this way BPAE is
in line with bare particularism as conceived by Baker (1967). On this view they are bare in that they have
no natures.
It is also in virtue of their lacking a nature that bare particulars cannot really be said to ground
the properties they contingently possess. In his defense of substrate theory, Mooreland (1998)
distinguishes between two different ways in which something might be said to have a property, one way
in which traditional substances are said to have properties and another way in which bare particulars are
said to have properties:
When a substance has a property, that property is seated within and, thus, an expression of the
inner nature of the substance itself. By contrast, bare particulars are simple and properties are
linked or tied to them (Mooreland 1998: 257).

64
Nelson (2007) makes use of this in confronting the apparent co-location (i.e. mereological identity) of a statue and
the lump of clay from which it is made. The idea is that both Satue and Lump are modes of one and the same
underlying bare particular.
65
Anscombe and Krner (1964) and Cambell (1990), among others, made this objection to bare particularism.

72

Properties do not simply depend upon substances in the sense that they are never to be found apart from a
substance, but also, and more importantly, because they are at least partially grounded in or caused by the
nature of the substances in which they inhere. The properties had by a bare particular, however, are
simply linked to them. One cannot help but wonder just what does the linking. Must the Platonic
demiurge link bare receptacles with Platonic forms or with otherwise free-floating tropes (i.e., property
instances)? According to Mooreland the substrate theorist takes such links and ties to be primitive.
Primitive is a term that we use relative to concepts but the fact that a bare particulars
properties are not grounded in it is not only a conceptual matter but also a metaphysical matter. There is
no problem if link here must be conceptually primitive, but it would be nice if we have a metaphysics
according to which the fact that objects possess contingent properties does not turn out to be some
mystery. One might at least explain property changes in terms of natural laws that hold between different
universals or trope classes, but now properties are doing all of the work and substrates find themselves in
danger of being sliced away by Occams Razor. Why not just conceive individuals as bundles of property
tropes or posit substansive natures?
A problem may also be raised against BPAE per se. BPAE is true to Quinean anti-essentialism in
that it is effectively part of deflationary account of des res modality:
Nothing in the anti-essentialist spirit runs contrary to assigning modal properties to objects
themselves, as long as those attributions are, ultimately, grounded in conceptual relations (Nelson
2007: 458).
In this respect BPAE is much like CCNC: the modal properties of objects are not grounded in
any of their actual properties or natures. The difference is that CCNC involves a strong form of
modalism, while for BPAE des res modality is simply a matter of how we conceptualize objects. The
modal properties of objects are not grounded in the objects themselves but in how we conceive of them.
Sometimes when we ask if x is possibly F we decide to hold some actual feature of x fixed, such
as its being of some sortal or of some material. According to BPAE, when we are not limiting our
consideration of an object to any description, every object turns out to be modally the same as every other
object. On such a view the meaning of modal utterances are determined by how we chose to describe or
conceptualize things, which is why everything becomes modally identical with everything else as soon as
we remove our contextual descriptions (i.e., as soon as we adopt the view from nowhere). But then, it
might be asked, why bother with SQML, or any form of QML in the first place? Would it not be easier to
leave the problems of QML to those who are foolish enough to believe that objects have essential
natures? To be sure, the Quinean will be happy to deny that there is such a thing as metaphysical
modality, but if we are looking to formulate a theory of metaphysical modality, BPAE is not for us.

73

Ordinarily, de res modality is supposed to concern what is necessary or possible for an object,
regardless of how we choose to describe it, at least when the modality at issue is metaphysical modality.
The non-deflationist regarding de res modality will believe that there is some fact of the matter as to
which properties of a given object are necessary to it and that such properties are typically grounded in
the nature or essence of that object. Hence, non-deflationary des res modality does require that objects
have nature, as Quine feared. But have we not already ruled out sortal essentialism? How can objects
have natures or essences if they do not necessarily belong to any given natural kind? First, we have only
denied essentialism relative to observable natural kinds (or at least those which have been observed by
humans) and second, objects might well have qualitative essences which are not natural kind essences.


















Copyright Adam Labecki 2013

74

CHAPTER FIVE:
TIMES, EVENTS, AND PROCESSES

The present chapter introduces temporality as well as events and processes. Regarding temporality as
such, it will be demonstrated that existence and identity are both permanent matters. This will mean that
the members D never come-into-being and never go-out-of-being. One consequence of this is that four-
dimensionalism will not be an option. Nevertheless, one can still speak of events and processes, the
occurrences of which are both contingent and temporary. Indeed, contingency will be a matter of which
events and processes occur and talk of contingent objects will ultimately be replaced by talk of contingent
processes of a special kind. Moreover, events and processes will be represented as structured entities
and, to my knowledge, the formalism to be presented will be unique in this respect.
66

1. Temporal Logic and the Permanence of Existence
Presumably anything that exists necessarily exists always exists. After all, it is intuitive to
suppose that if something exists necessarily then it must exist unconditionally. Williamson (2002)
mentions this matter in passing, assuming that necessary existents exist at all times.
67
Given that
metaphysical modality, as understood in this text, is considerably different from mere conceptual
modality, however, we should not rely too strongly on our intuitions. Fortunately, it is well known that
the simplest quantified temporal logic renders objects eternal just as easily as SQML renders them
necessary.
Tense logic, beginning with the pioneering work of Prior (1957, 1968), is largely motivated by
the intuition that only the present genuinely exists. As Augustine famously wrote, the past is no longer
and the future not yet is. The doctrine that only the present really exists is known as presentism and it
is effectively equivalent to endorsing the A-theory of time. Many metaphysicians, however, regard time
as literally a fourth dimension. On this view the past and future exist just as much as the present; their
difference from the present being no more special than the way there differs from here. Paris is not
spatially present to me, but this gives me no reason, on pains of solipsism, to deny that it exists. Likewise,
429 A.D. is merely behind us and 3500 A.D. is merely ahead of us. The doctrine that the past and
future are just as real as the present is known as eternalism and it is effectively equivalent to endorsing
the B-theory of time.

66
Kims analysis, upon which mine is partially based, is also structured, but he does not provide a logic of events,
much less a quantified logic of both singular and plural events.
67
Williamson (2013), however, recognizes that necessitism could hold while permanentism fails to hold, but
suggests that most necessitists will also be permantentists.

75

According to eternalists, tense is not metaphysically fundamental. While tense logic may be
indispensable in the formal analysis of temporal language (and our first-person conception of time), the
eternalist will not make serious use of it when doing metaphysics. The eternalist, however, does not deny
that there is such a thing as time
68
and hence should still have interest in temporal logics that do not
appeal to tense.
I shall briefly discuss classical quantified tense logic simply in order to demonstrate that it
verifies permanentism. Tense operators work much the way as modal operators both are effectively
devices for indexing Tarski models (only in this case the set of said models really must be ordered relative
to a frame). The weakest form of temporal logic is known as Minimal Tense Logic K
t
, so named because
is the temporal equivalent to the modal system K.
Sketch of K
t
:
Primitive Terms:
H: It has always been the case that
G: It will always be the case that
Defined Terms:
P H It was the case that
F G It will be the case that
( ^ (H ^ G)) It is always the case that
( \ (P \ F)) It is at some time that case that
Axioms and Rule:
TL1: H( ) (H H)
TL2: G( ) (G G)
TL3: HF
TL4: GP
ETR: If , then H and G

68
Whether eternalism amounts to the elimination of our ordinary conception of time is quite another question and
one on which I have no strong opinion (it should be noted that Augustines analysis of time was phenomenological
in character). What is quite certain, however, is that it does not amount to the elimination of our scientific
conception of time, which is relatively flexible.

76

H and G are both analogous to the necessity operator, where H ranges over the entire past and G ranges
over the entire future. They are often referred to as the strong temporal quantifiers. Likewise, P and F
are analogous to the possibility operator, where P rangers over some past time and F rangers over some
future time. They are often referred to as the weak temporal quantifiers and are defined in terms of H
and G respectively (although and are more tightly analogous to . and ).
TL1 and TL2 are direct analogs of the modal K axioms while TL3 and TL4 ensure that the strong
and weak operators are interdefinable. Finally ETR, is short for eternalization is the temporal analogue
of the rule of necessitation. Temporal logic requires the addition of two new model elements: a non-
empty set T and a binary relation <. Intuitively speaking, T is the set of all moments of time while < is
the before relation.
69

Expressions that are indexed according to both worlds and times are effectively two-
dimensional or multi-modal. When both modalities are in effect, the intensions of n-place predicate are
treated as n+2 tuples, indexed to both a world w and a time t. Accordingly, the truth valuation for atomic
formulas would be modified simply by adding the temporal index:
[Vn
t
] V
M
(r
1

n
, w, t) = 1 iff ([o
1
]
M
, .... , [o
n
]
M
, w, t) e I(r)
The other valuation clauses would be modified as expected. The truth valuation clauses for the temporal
operators are as follow:
[VH] V
M
(H, w, t) = 1 iff for every t such that t < t, V
M
(, w, t) = 1
[VG] V
M
(G, w, t) = 1 iff for every t such that t < t, V
M
(, w, t) = 1
Within this minimal tense logic it is easy to show that all objects permanently exist.
T*: Permanence of Existence (PE): !xy(x = y)
1. y x=y CQT
2. Gy x=y 1. ETR
3. Fy x=y 1. ETR
4. y x=y 1-3 Def.
5. !xy x=y 4. UG

69
K
t
is neutral regarding a number of questions regarding T. It is indifferent as to whether or not T is dense, or
discrete, has a minimal element or a maximal element etc. Indeed, K
t
alone does not even guarantee that < is a
partial ordering over T, much less a total ordering. When analyzing events it will be presumed that time is not dense.

77

One will have noticed that the proof is effectively the same as the proof for NE. Combining temporality
with ordinary modality, one can easily derive !x.y(x=y) by applying the rule of necessitation
between lines 4 and 5.
The official proof for PE, however, will not make use of tense logic as such, but will involve
direct quantification over times. This decision is made for two reasons. First, the logic of events will be
more expressive if quantification over times is permitted. Second, I shall ultimately need to adopt
eternalism, and so there is not much need for tense operators (although the eternalist operators by
themselves would have been fine).
Let t, with or without subscripts, serve as the variable for times or moments. Non-italicized t, with
or without subscripts will be used on the rare occasion in which time constants might be required.
How much metaphysical weight should be placed upon these times or moments? Not too much
and not too little! Times are semantically indexical devises, much like worlds. Now, of course, they are to
be intrepreted as representing something or other and in the very least they represent successive states of
a given world.
The set T of times will be left neutral in several respects and hence it will not say that much
about the nature of time. As for those features of time which will be decided upon, I do not want to insist
that all of them characterize the nature of time as such. For instance, if time is interdefinable with space in
such a way that there is no such thing as absolute simultaneity, then quantification over times will be
misleading.
Should we be very doubtful regarding absolute simultaneity? While there is no absolutely
simultaneity between events in the spacetime of Special Relativity (SR), there are a number of
scientifically motivated reasons for supposing some absolute time-frame:
1. QM assumes an absolute time-frame for the evolution of the wavefunction.
2. In cosmological applications of General Relativity (GR), which effectively supersedes SR,
absolute simultaneity re-emerges. For instance, statements about the age of the universe are made
relative to a privileged cosmological time-frame (Swinburne 2007).
3. Relative simultaneity might well be experimentally challenged by observations of Bells
inequalities, such as Aspect et. al. (1982). This is because wave collapse of a physical system must be
simultaneous for each particle of the system, regardless of the distance between them.

78

Accordingly, the wide-spread belief that SR did away with absolute simultaneity once and for all is
misguided.
70
From considerations of total physics the case for absolute simultaneity looks stronger than
the case against it. If, however, SR is right about absolute simultaneity, then any interpretation will have
to be understood to be relative to an arbitrarily chosen timeframe.
Time will also be construed as discrete. This condition will later be important when analyzing
modality in terms of diverging worlds, but I also believe that time is probably discrete. When I stake
my bets that time is discrete I affirm the following disjunction: if there is such a thing as spacetime at all,
it is made up of discrete Plank-scale units or, if there is no such thing as spacetime, there is a minimal
amount of time during which a physical event can be said to take place. In each case the unit itself might
be contingent, but there would necessarily be some unit or other. Again, the supposition of continuous
spacetime leads to infinities in QFT and the infinities that arise from gravitational fields cannot even be
renormalized without quantizing gravity itself, and hence spacetime. If string theory is true, and all
intrinsic properties are string-vibrations or oscillations, then it would be meaningless to speak of
properties being possessed at some literal point of time, since no oscillation will occur during such an
instant. Indeed, this is a lesson one might have already learned from Zenos arrow argument. If I
happen to be mistaken, and time is genuinely continuous, then, once again, the unit here must be
interpreted as an arbitrary but appropriate choice, perhaps Plank time would do.
Nevertheless, I do not want to make too much of the times or moments of the temporal logic.
Their primary purpose is to relativize predication to a linear sequence of moments. This is partly
because I am uncertain that time is really fundamental at all. It might well be the case that spatiotemporal
relations can be reduced to ordered causal relations, as proposed by the Causal Sets theory of quantum
gravity.
In terms of model theory, then, we shall introduce a non-empty set T of times (or moments) and a
relation < on T requiring that T should be linear and discrete. Axiomatically speaking, quantification over
times will be just like quantification over objects. It will be necessary to update the logic of identity so as
to accommodate times:
I1: =
t
(Law of Identity)
I2: =
t
!t((o, t) (, t)) (Indiscernibility of Identicals)
The t subscript is added to denote that o is identical with itself, or , at t. The < relation will also be
introduced into the object language. The following axioms will govern <:

70
See Craig and Smith (eds.) (2007) for a collection of papers defending absolute simultaneity in the context of
contemporary physics.

79

T1: !t t < t (Time Irreflexivity)
T2: !t
1
!t
2
!t
3
((t
1
< t
2
^ t
2
< t
3
) t
1
< t
3
) (Time Transitivity)
T3: !t
1
!t
2
((t
1
< t
2
\ t
2
< t
1
) \ t
1
= t
2
) (Linear Ordering)
T4a: !t
1
!t
2
(t
1
< t
2
t
3
(t
1
< t
3
^ t
4
t
1
< t
4
< t
3
)) (Discreteness-a)
T4b: !t
1
!t
2
(t
1
< t
2
t
3
(t
3
< t
2
^ t
4
t
3
< t
4
< t
2
)) (Discreteness-b)
T5: t
1
< t
2
. t
1
< t
2
(Necessity of Priority)
This sort of language is, in many respects, more expressive than tense logic. For instance, one can express
there is a beginning of time and there is no beginning of time as follows:
t
1
t
2
t
2
< t
1

!t
1
t
2
t
2
< t
1

One can similarly express the propositions there is an end of time and there is no end of time.
t
1
t
2
t
1
< t
2

!t
1
t
2
t
1
< t
2
Given discreteness, we can also define a successor function between times.
Df16: Successor: s(t
1
) = t
2
(t
1
< t
2
^ t
3
t
1
< t
3
< t
2
)
The inverse of the successor function will get us the predecessor of a time:
Df17: Predecessor: p(t
1
) = t
2
s(t
2
) = t
1

Having introduced times, ordinary first-order formulas will officially be replaced by their
temporal counterparts. That is to say, Fxy will now be understood as an elliptical formula requiring
that a time be added to the relata: Fxyt, which is to be read as x is F-related to y at t. More precisely,
where r
n
is an n-place predicate and o
1
o
n
and n-tuple of objects, r
n
o
1
o
n
t is a temporal atomic
formula, or a simple event. If Fxt
1
and Gxyt
2
are both true and t
1
< t
2
we may also say that Fxt
1

< Gxyt
2
.
71
That is, events inherit temporal order from their time-relata. The primary difference, however
is that the ordering of events is not linear, since more than one event may occur at a given time. In the
treatment of events below a simultaneity operator will be introduced to indicate events which occur at the
same time.
Some might complain that this way of temporally relativizing relation instances effectively
renders each n-place relation as an n+1 place relation. Being taller than is no longer a two place
relation, but a three place relation, where the first two relata are objects and the third is a moment of time.
Nevertheless, it might still be regarded as a two-place relation precisely because it takes two objects.
After all, in the standard semantics for SQML relations are already relative to worlds, the only difference
is that we do not include world variables within the object language (since modal operators are used

71
Although in the latter case < is effectively an operator rather than a relation.

80

rather than direct quantification over worlds). Moreover, I might have decided that times are directly
predicated by abstracts formed from first-order atomic formulas. Where [ ] is an abstraction operator that
turns a formula into a complex term, [Fxy] would be a complex predicate, which might be regarded as a
state of affairs. Such complex terms can be attributed to times such that [Fxy]t would read something
like: x bearing F to y is instantiated at t. While in some respects I prefer this approach, it would require
additional syntactical terms, namely a variable for states of affairs, and the abstraction operator.
2. Permanence of Existence and the Denial of Four-Dimensionalism
T19: Permanence of Existence (PE): !x!ty x =
t
y
1. !y x
t
y x
t
x CQT1
2. x =
t
x !y x
t
y 1. PL
3. x =
t
x I1
4. !y x
t
y 2,3. MP
5. y x =
t
y 4. QI2
6. !x!ty x =
t
y 5. UG
Corollary: !x.!ty x =
t
y
No object ever comes into existence or goes out of existence there is no becoming with respect to
existence. The corollary effectively combines PE with NE: everything exist in every possible world at
every time. The temporal UG on line 6 is clearly justified, since the initial t is perfectly general (i.e., t was
not introduced by assumption). When something or other is both necessary and permanent I shall say that
it is fixed. Hence we may now speak of the fixity of existence across all w, t pairs.
PE is suggestive of Leibnizs doctrine that substances have no beginning or end in time but,
rather, undergo transformations.
We must then know that the machines of nature have a truly infinite number of organs, and are so
well supplied and so resistant to all accidents that it is not possible to destroy them. A natural
machine still remains a machine in its least parts, and moreover, it always remains the same
machine that it has been, being merely transformed through the different enfolding it undergoes,
sometimes extended, sometimes compressed and concentrated, as it were, when it is thought to
have perished (Leibniz 1989 (1695): 142).
Indeed, in Chapter Three something along these lines was alluded to when speaking of the modes of
necessary bare particulars. At present it would be premature to begin speculating on how a new theory of
transformations might be construed, since there are other ontological questions, yet to be addressed,
which may well preclude the need for such a theory. Perhaps there are no such things as living beings at
all and hence nothing for a theory of transformations to explain. Likewise, if existence monism should
turn out to be true, then we have at most one animal that would undergo such transformations.

81

Df19: Endurent: For every x, x is an endurent if and only if:
t
1
t
2
(t
1
t
2
^ y(x =
t1
y ^ x =
t2
y)
An endurent is an object that strictly exists over time, relative to which one may speak of identity over
time.
Do14: Endurentism is the doctrine that some objects are endurents.
Do15: Perdurantism is the doctrine that there are no endurents.
Do16: Universal Endurentism is the doctrine that all objects are endurents.
T20: Permanence of Identity (PI): !x!yt
1
(x =
t1
y !t
2
x =
t2
y)
1. !t
2
x =
t2
y I1. UG
2. x =
t1
y !t
2
x =
t2
y 1. PL
3. t
1
(x =
t1
y !t
2
x =
t2
y) 2. EI
4. !x!yt
1
(x =
t1
y !t
2
x =
t2
y) 3. UG
Corollary: Universal Endurentism is affirmed unless |T| = 1.
Corollary: Perdurantism is denied unless |T| = 1.
At first one might have supposed PI would directly entail both universal endurentism and the denial of
perdurantism, but perdurantism would be vacuously true if T contains but a single member. Nevertheless,
perdurantism is just as well as defeated by T22, since it can only be true if the universe is practically
timeless. That is, to say that there is but a single moment of time is just as much as to say that time is an
illusion.
Endurentism as such may seem to be obviously true, but it happens to be controversial in the
contemporary literature on the metaphysics of material objects. Recall that Quine (1950), Lewis (1986a),
Armstrong (1997), and Sider (2001) endorse the doctrine of supersubstantialism, according to which the
fundamental objects are all spacetime points. By PI, however, we have seen that spacetime points cannot
be included in D.
The controversy over endurentism arises due to several paradoxes regarding change, where two
alternatives to endurentism, namely perdurantism and exdurantism respond to said paradoxes by denying
that there is any such thing as strict identity over time. Given PI, retreating from endurentism will not be
an option when responding to such paradoxes. The problem of temporary intrinsic properties will be
discussed in Chapter Nine.

82

Basic Mereology
72

P: Part: Pxy is read x is a part of y.
Part is a primitive term.
M1: Transitivity of Parthood: (Pxy ^ Pyz) Pxz
M2: Reflexivity: Pxx
M3: Antisymmetry: (Pxy ^ Pyx) x = y
Df19: Proper Part: x is a proper part of y, written PPxy, if and only if:
Pxy ^ x = y
Df20: Overlap: x overlaps y, written Oxy if and only if:
z(Pzx ^ Pzy)
M4: Strong Supplementation: Pyx z(Pzy Ozx)
Df21: Disjoint: x and y are disjoint, written Dxy if and only if:
Oxy
Df22: Underlap: x and y underlap, written Uxy if and only if:
z(Pxz ^ Pyz)
Do17: Presentism: !t
1
t
1
= t
@
(Where t
@
is the present moment, by analogy with the actual world)
Do18: Eternalism: !t
1
t
2
t
1
= t
2

T21*: Eternalism: !t
1
t
2
t
1
= t
2

!t
1
t
2
t
1
= t
2
is a simple theorem of CQT when quantifying over times. But using it as a refutation of
presentism is perhaps a little too fast. The presentist might insist that classical quantification over times is
permissible provided that we are talking about ersatz times. After all, had we decided to quantify over
worlds themselves, we would not immediately commit ourselves to possibilism (i.e. modal realism).
Hence I have placed an asterisk to note that the eternalism here need not be taken to exclude
presentism. Presentism, however, will surely be in trouble if past or future events can serve as truth-
makers for either tensed or non-tensed formulas. It will also be in trouble if there is a fact of the matter as
to which events actually occur not only in the past, but also in the future.

72
Contemporary mereology has its origins in Leniewski (1916) and Lenard and Goodman (1940).

83

The following definition of instantaneous temporal part is taken from Sider (2001):
Df23: Instantaneous Temporal Part: x is an instantaneous temporal part of y at time t
1
if and
only if:
Pxyt
1
^ (!t
2
(E!xt
2
t
2
= t
1
) ^ !z(Pzyt
1
Oxzt
1
))
In words: x is an instantaneous temporal part of y at time t
1
just in case, x is a part of y at t
1
and for every
time t
2
, if x exists at t
2
, then t
2
is identical with t
1
and for every z if z is a part of y at t
1
, then x overlaps z at
time t
1
. This definition is intended to capture the idea that each temporal part of an object is a temporal
slice of some purdeurnt object.
T22: There are no instantaneous temporal parts unless for every t
1
, t
2
e T, t
1
= t
2
(i.e., unless
|T| = 1)
This is an immediate consequence of PE and the definition of instantaneous temporal part. Every object
exists at every time, hence there can be instantaneous objects only if there is but a single instant of time.
It is quite questionable whether it is even possible for there to be but a single moment of time.
That is, a model in which there is only a single member of T does not seem to differ from a model a world
that is eternal in the sense in which there is no time at all: there is but a single unqualified state of the
world. Moreover, it is quite questionable whether time could come to an end in a naturalistic manner. At
any rate, unless time is wholly illusionary, there cannot be any instantaneous temporal parts.
Do19: Four-Dimensionalism (4D) is the doctrine that (1) Eternalism is true and (2) all objects
are either temporal parts or composed of temporal parts.
T23: Four-dimensionalism fails on any model in which there is more than one member of
T.
This is an immediate result of the definition of 4D and T24.
No possible world is an interesting four-dimensional world, since 4D can only obtain in worlds that are
practically timeless.
Do20: Spacetime Substantialism is the doctrine that spacetime points are first-order objects.
Spacetime points, no matter how they might be defined, are going to be temporary existents and so:
T24: Spacetime Substantialism is false unless |T| = 1.
Perhaps I have been too quick to conditionally refute 4D. Perhaps 4D theorists need not construe temporal
parts as first-order-objects. After all, on 4D, objects are construed rather more like occurents. That is,
perhaps quantification over temporal parts is really quantification over events while fusions of temporal

84

parts are like processes. Events and processes will be discussed below and have first-order objects as
constituents and hence cannot exist without such objects. This will not help matters for the 4D theorists,
since the higher-order items are going to have to be built upon the basis of permanent first-order objects,
while 4D denies that there are any such things.
3. Events and Processes
Now that temporality has been introduced, events may be introduced by rendering each n-place relation as
an n+1 relation, where the extra relata is a member of T. The result is similar to the manner in which Kim
(1973, 1976) treats events. It is arguably the case that atomic first-order formulas implicitly involve a
temporal relata or indexes in the first place. What do we mean when we simply say x is F? Sometimes we
implicitly refer to the time of the utterance, so that x is F is construed as x is F now. This is not,
however, the case in such propositions as Socrates is snub-nosed. Hence, in general, Fx might be
construed as t Fxt.
Many may find the Kim-style an analysis of event to be objectionable but I do not intend to be
giving a general analysis of how the word event behaves in ordinary language. Nor am I trying to
provide an analysis of our concept of event. I basically want to be able to talk about predicates being
instantiated by objects at times and I am naming such instantiations events. An event need not even
indicate that any change has occurred. For instance, suppose t
1
= p(t
2
) and both Fxyt
1
and Fxyt
2
are
true. Fxyt
2
names an event no less than Fxyt
1
.
I will, however, note that, some the most glaring difficulties that arise on Kims analysis
73
will be
removable once the notion of natural property is officially introduced. Moreover, Kims approach to
analyzing events is by far the easiest to logically model.
74
It will, however, be useful, to model not only
simple events, involving a single predicate instantiation, but also events involving several such
instantiations occurring at a given time. In addition to these, we will eventually want to model processes,
or events which unfold over a span of time.
A simple event is effectively an atomic formula with the addition of a time, i.e., r
n
o
1
o
n
t. I
shall use e, with or without subscripts, as the variable for simple events. While event variables count as
formulas, it will be useful to assign them a limited kind of quasi-predication, which I shall call

73
In particular, I have in mind that, on Kims analysis, Brutus stabbing Caesar and Brutus stabling Caesar
violently come out as two separate events when it seem that they are really just two distinct ways of describing one
and the same event.
74
The more sophisticated analysis of events found in Davidson (1967, 1969) is focused upon action verbs, instances
of which are analyzed as three-place relations involving an agent, a patient, an event. Hence events themselves are
provided with a primitive domain.

85

involvement. If e is in event variable substituting for X
n
x
1
.x
n
t, then e can be said to involve X
n
,
x
1
.x
n
, and t: (eX
n
^ ex
1
x
n
^ et. Involvement is syntactically distinct from ordinary
predication in that et does not entail e. Moreover, there will be some additional logically defined
predicates that may be assigned to event variables. Extra-logical predicates (i.e., those which are not
logically defined), however, can never be predicted of an event.
Quantification over events is not quantification into name position but quantification into formula
position and in this respect it is similar to quantification over propositions. Just as X
n
x
1
.x
n
t can stand
alone as a formula, so can e, since one can substitute the one for the other. Again, while events will
exist necessarily, they only occur (or take place) in some worlds. This is because e e is a
perfectly coherent formula.
It must be admitted that treating events in this way comes at some cost. Non-occurring events
would appear to be abstract objects, while occurent events do not seem to be abstract, and hence concrete.
It should be noted, however, that the object language does not attribute such predicates to events.
Nevertheless, it must be admitted that, meta-linguistically speaking, non-occurent events would be
construed as abstract objects while occurring events are, in some lose sense, concrete. If this is so, am I
not treating occurent events as contingently concrete objects and non-occurent events as contingently
abstract objects? Again, strictly speaking no, since in the strict sense of the term events are not the sort of
things that can really be abstract or concrete; there is something of a category mistake at work here. More
importantly, the fact that simple events, being equivalent to temporalized atomic formulas, are naturally
denoted by formulas rather than terms means that the distinction between existent and occurent events is a
natural result of logic itself.
75
I do, however, admit something like contingent concreteness and
abstractness is being attributed to events and that it is of some concern. The primary appeals here is that
our intuitions about the metaphysics of events are not terribly well-formed in the first place and that
treating events in this way is a direct result of logical considerations and not of some additional
theoretical decision, as in the case of positing contingently non-concrete objects.
Also, some defenders of four-dimensionalism may wonder why they could not help themselves to
this distinction when quantifying over temporal parts. My response is that they could, but not in the
context of a first-orderizable theory, where first-order quantification is certainly over terms admitting of
non-logically defined predication. So four-dimensionalism might be salvaged within a fundamentally
propositional language, but this comes at a considerable cost to expressiveness.
One decision that must be made in the formation of any theory of events (or facts) is whether
events that necessarily obtain in the same possible worlds are identical. If so, then we can replace identity
with the necessitated bi-conditional .(X
n
x
1
x
n
t e). This is sometimes referred to as logical

75
This would still have been the case if events were construed as state of affairs instantiated by times.

86

identity. While this would be convenient, I am not convinced that necessary co-occurrence must yield
identity For instance, if non-identical a and b are essentially F and G respectively, then, for any t, .(Fa
Gb) will be true. Perhaps more importantly, logical identity raises the threat of the dreaded Slingshot
argument. Events are fact-like objects and so they are potentially threated by the Slingshot. One of the
standard premises of the Slingshot argument is that facts which are logically equivalent are identical.
Denying this, however, requires that we discern another standard of identity for events. This is one place
where the introduction of structure (involvement) comes in handy: two events are identical just in case
they involve all the same items.
Rather than introduce separate variables for compound events and processes, I shall introduce a
plural event variable: ee (with or without subscripts). Plural events in general are utterly gerrymandered
but various sorts of occurent objects can be obtained by placing various restrictions on which sort of
atomic events may be contained in a plural event.
Two events will be said to be simultaneous, written e e just in case for any t e T, et if and
only if et. A compound event, written Eee, is just a plural event ee such that any two simple events
contained in ee are simultaneous. Containment will be written e ee or X
n
x
1
.x
n
t ee and is
formally like the membership relation of set-theory. Furthermore, a plural event involves anything that is
involved by any simple event that it contains.
Let a span, written t
m
t
n
, be a set of linearly ordered discrete times (or moments) without
gaps (i.e., any t
i
that is after t
m
and before t
n
it is a member of t
m
t
n
). A thin process is a sequence of
simple events along a span of time: e
m
t
m
e
n
t
n
. As shorthand, one may leave out the involved times
and just write [e
m
e
n
]. Likewise, a thick process in general is a sequence of compound events:
ee
m
ee
n
]. P and p will be used as predicates for thick process and thin process respectively.
The difference between a thin process and a compound event is that no two simple events
contained in a thin process involve the same time while every simple event contained in a compound
event involves the same time. All of the definitions mentioned in the last few paragraphs (and others
besides) will be clearly laid out below.
4: The Logic of Times, Events, and Processes
New Primitive Terms:
1. Time Variables: t
1
, t
2
.
2. Singular Event variables: e
1
, e
2
.
3. Plural Event variables: ee
1
, ee
2
.
New Primitive Connectives:

87

, <, =
(Using identity for formulas is technically an operator as opposed to the already introduced
identity relation)
Expanded Grammar:
1. If r
n
is an n-place predicate, where n 0,
1

n
are terms, and t is a time, then r
1

n
t is
a formula (namely an atomic formula).
2. If and both denote events then = if a formula
3. If denotes a singular event and denotes a plural event, then is a formula.
Defined Terms and Relations:
Df24: e-Involvement: e ((e = r
n
o
1
o
n
t) ^ (( = r
n
\ = o
1
o
n
) \ =
t))
Df25: e-Monadic: Me er
1

Df26: Inclusion: (ee
1
ee
2
) !e(e ee
1
e ee
2
)
Inclusion is to containment what the subset relation is to the membership relation.
Df27: Span: St
1
t
n
!t
i
(t
i
< t
i
< t
n
(t
i
t
1
t
n
))
Df28: Thin Process Constant: [e
1
e
n
] t
1
t
n
(St
1
t
n
^ (e
1
t
1
^ e
2
t
2
^ e
n
t
n
))
[e
1
e
n
] may be used to denote a sequence of simple events ordered along a time span.
Df29: ee-(Weak) Involvement: ee e((e ee) ^ e)
Df30: ee-Strong Involvement: ee[] !e((e ee) ^ e)
Df31: Compound Event: Eee t(ee[t])
A compound event is simply a plurality of events that strongly involves a particular time.
Df32: Thick Process Constant: [ee
1
ee
n
] t
1
t
n
(St
1
t
n
^ (ee
1
[t
1
] ^ ee
2
[t
2
] ^
ee
n
[t
n
]))

Df33: Event Priority: (e
1
< e
2
) ((e
1
t
1
^ e
2
t
2
) ^ t
1
< t
2
)
Df34: Event Simultaneity: (e
1
e
2
) ((e
1
< e
2
) ^ (e
2
< e
1
)
Definitions 34-38 are to be understood relative to compound events.

Df35: Strong-ee Temporal Priority: (ee
1
<
S
ee
2
) !e
1
!e
2
(((e
1
ee
1
) ^ ((e
2
ee
2
) e
1
< e
2
)))
One plurality of events is strongly prior to a second just in case every event contained in the first is
temporally prior to every event contained in the second.


Df36: ee-Weak Temporal Priority: (ee
1
<W ee
2
) e
1
((e
1
ee
1
) ^ !e
2
((e
2
ee
2
) (e
1
< e
2
)))

88

One plurality of events is weakly prior to a second just in case at least one event contained in the first is
prior to every event contained in the second.

Df37: ee-Temporal Overlap: (ee
1
<
O
ee
2
) ((ee
1
<
S
ee
2
) ^ (ee
2
<
S
ee
1
))
Plural events ee
1
and ee
2
temporally overlap just in case there is some time t such that (e
1
ee
1
) ^ (e
2

ee
2
) and e
1
t ^ e
2
t.
Df38: ee-Temporal Underlap: (ee
1
<
U
ee
2
) ((ee
1
<W ee
2
) ^ e
1
((e
1
ee
1
) ^ !e
2
((e
2
ee
2
)
(e
2
< e
1
)))
One plural event temporally underlaps a second just in case the earliest member of the first is earlier than
the earliest member of the second and the latest member of the first is later than the latest member of the
second.

Df39: ee-Temporal Simultaneity: (ee
1
ee
2
) ((ee
1
<W ee
2
) ^ (ee
2
<W ee
1
))
Two plural events are effectively simultaneous just in case the earliest members of each occur at the same
time and it the latest members of each occurs at the same time.
Df40: Process: Pee
1
t
1
t
n
(St
1
t
n
^ !ee
2
((ee
2
ee
1
) (ee
2
[t
1
]\ ee
2
[t
n
])))
Df41: Thin Process: pee (Pee ^ !t!e
1
!e
2
((((e
1
ee) ^ e
1
t) ^ (((e
2
ee) ^ e
2
t)))
e
1
= e
2
))
Df42: p-Stage: (e
t
ee) ((e ee) ^ pee)
Df43: P-Stage: (ee
1

t
ee
2
) ((ee
1
ee
2
) ^ (Pee
2
^ Eee
1
))
Stages are like instantaneous temporal parts of processes.
Df44: Sub-Process: (ee
1

P
ee
2
) ((ee
1
ee
2
) ^ (Pee
1
^ Pee
2
))
Df45: Segment: (ee
1

S
ee
2
) ((ee
1

P
ee
2
) ^ ((Pee
2
^ Pee
1
) ^ (ee
2
<
U
ee
1
)))
Df46: Embedded-Process: (ee
1

E
ee
2
)

((ee
1

P
ee
2
) ^ (ee
1
ee
2
))
Axioms
The standard axioms for classical quantification and identity are extended to the treatment of times,
events, and plural events, although the extension of CQT1 to events requires some spelling out due to the
complex nature of event formulas.
Temporal Logic:
t1: t < t (t-Irreflexivity)
t2: (t
1
< t
2
^ t
2
< t
3
) t
1
< t
3
(t-Transitivity)
t3: (t
1
< t
2
\ t
2
< t
1
) \ t
1
= t
2
(t-Linearity)

89

t4a: t
1
< t
2
t
3
(t
1
< t
3
^ t
4
t
1
< t
4
< t
3
) (t-Discreteness (a))
t4b: t
1
< t
2
t
3
(t
3
< t
2
^ t
4
t
3
< t
4
< t
2
) (t-Discreteness (b)
t5: t
1
< t
2
. t
1
< t
2
(Necessity of Priority)
Event Logic:
e1: !e(e) (ro
1
o
n
t/e) (e-CQT1)
Derived: e-Existential Introduction (ro
1
o
n
t) e(e/ro
1
o
n
t)
ee1: !ee(ee) ((e
1
^e
n
)/ee) (ee-CQT1)
ee2: e(e) ee!e((e ee) (e)) (ee-Comprehension)
ee3: !eee(e ee) (ee-Non-Vacuity)
ee4: !e((e ee) (e ee)) (ee = ee) (ee-Identity)
ee4: (e ee) .(e ee) (Nec, of Containment)
ee5: ee !e((e ee) e) (ee-Occurrence)
Pee1: ee((ee ^ (pee ^ ee[]) ee((Pee ^ (ee ee)) ^ !e(e ee (e ^ e))))
(Pee-Comprehension)
The logic of plural events is, of course, very similar to plural logic for terms except for ee5, which arises
on the basis of the fact that events variables are formulas rather than terms (ee5 is something like a
principle of analytic entailment see van Fraassen (1969)). Pee1 is required to specify processes
centered around the inclusion of a specific n-tuple of objects or property. If there is a thin process ee that
strongly involves then there is a (typically properly-thick) process ee that contains every -involving
event that occurs during the span of ee.
Semantics
Definition of a Model
A model, ML , is five-tuple ML = (D, W, T, <, I,) such that:
D is a non-empty set of objects: the domain
W is a non-empty set of worlds
T is a non-empty set of times
< is a relation on T satisfying the following conditions:
(1) T is a strict linear ordering:

90

(a) !t t < t
(b) !t
1
!t
2
!t
3
((t
1
< t
2
^ t
2
< t
3
) t
1
< t
3
)
(c) !t
1
!t
2
(t
1
< t
2
\ t
2
< t
1
\ t
1
= t
2
)
(2) For any t
1,
t
2
e T, if t
1
< t
2
, then there exists some t
3
e T such that t
1
< t
3
and
there is no t
4
e T such that t
1
< t
4
and t
4
< t
3
(i.e., T is discrete).
I is the interpretation function satisfying the following conditions:
(a) if o is an object constant then I(o) e D
(b) if t is a time constant then I (t) e T
(b) if r
n
is an n-place predicate constant then I(r
n
) assigns r
n
to a set of n+2 tuples x
1
,
,x
n
, t, w
76

Let be a set such that, for each positive integer n, if r
n
is an n-place predicate, I(o
1
), , I(o
n
)
e D
n
, and t e T, then I( o
1
), , I(o
n
), I(r
n
)

, I (t) e . is the set of all events, possible or
otherwise.
Let be the variable assignment subject to the following conditions:
1. For each time variable t, [t]
M
e T
2. For each simple event variable e, [e]
M
e
3. For each plural event variable ee, [ee]
M ,
The valuation-function V
M
for ML and a variable assignment , is a function that assigns 1 or 0 (true or
false) to each formula relative to each world according to the following conditions:
[V7
t
] V
M
(!t, w) = 1 iff for every t e T, V
M(t/t)
(, w) = 1
[Vn
t
] V
M
(r
1

n
t, w) = 1 iff [
1
]
M
, [
n
]
M,
[t]
M
e I (r)
(w)
77


[V.] V
M(., w) = 1 iff for every w, t, w e W and t e T, V

((t), w) = 1
[V7
e
] V
M
(!e, w) = 1 iff for every e e , V
M(e/e)
(, w) = 1
[V
e=
] V
M
(e = r
1

n
t, w) = 1 iff [e]
M
= r
n
, [
1
]
M
, [
n
]
M
,

[t]
M
[Ve] V
M
(e, w) = 1 iff [e]
M
= r
n
, [
1
]
M
, [
n
]
M
,

[t]
M and [
1
]
M
, [
n
]
M
, [t]
M
e
I (r
n
)
(w)

[V7
ee
] V
M
(!ee, w) = 1 iff for every ee , , V
M(ee/ee)
(, w) = 1
[V] V
M
(e ee) iff [e]
M
e [ee]
M

[Vee] V
M
(ee, w) = 1 iff for each e such that e ee, V
M
(e, w) = 1

76
Alternatively, it could be said that I assigns each p to a function g: TW (D
n
).
77
Where I (r)
(w)
is understood to be the interpretation of r restricted to a world w (i.e., I (r)
(w)
includes all and only
those members of I (r) of which w is a member.

91

Compound events, thin processes, and thick processes, might have been provided with their own variables
and quantifiers by defining the respective derived domains. These definitions will be useful for keeping
track of the semantics of some of the more sophisticated formulas of the object language.
1. Let
E
a subset of () such that S e
E
if and only if:
For any s
1
, s
2
e , if s
1,
s
2
e S then, for any time t, t e s
1
if and only if t e s
2
. That is,
E
is the
set of all possible compound events.
2. Where T is the set of all times, let C be a subset of (T) subject to the following condition:
Where B e (T), B e C if and only if, for all t
1,
t
2
such that t
1
< t
2
,

t
1
e B and t
2
e B implies
that for any t
3
, if t
1
< t
3
<t
2
, t
3
e B. Roughly speaking, C is the set of convex subsets of T,
78
that is,
each member of C is a time-span.
3. Let
p
be a set of subsets of {
1
,
2
,
n
} such that, for any positive integer n, s
n
e
p
if and only
if:
There exists some B e C such that B= n and for any t, if t e B there is exactly one s e s
n

such that t e s.
4. Let
P
be a set of subsets of {(
1
), (
2
), (
n
)} such that, for any positive integer n, S
n
e
P
if
and only if:
There exists some B e C such that B= n and for each t e T, t e B there is some S , S
n
such
that each s e S is such that t e s.
5. Some Event Logic Theorems
T25: !x!tX Xxt
1. !x!ty x =
t
y PE
2. y x =
t
y 1. CQT
3. E!xt 2. Def. E!
4. !x!tX Xxt 3.CQT
Every object always has some property or other (i.e., namely existence).
T26: e-Comprehension: !X
n
!x
1
x
n
!te.(X
n
x
1
x
n
t = e)
1. !ee.(e = e) e.(X
n
x
1
x
n
t = e) e1
2. !ee.(e = e) NE
3. e.(X
n
x
1
x
n
t = e) 1-2. PL
4. !X
n
!x
1
x
n
!te.(X
n
x
1
x
n
t = e) 3. UG

78
Although convex is typically reserved for real-valued sets.

92

For every n-place predicate X
n
, n-tuple of objects, x
1
x
n
and time t, there is some event that is identical
to x
1
x
n
instantiating X
n
at t. This is something like a comprehension principle for simple events.
T27: !e((Xx
1
x
n
t = e) .(Xx
1
x
n
t e))
1. (Xx
1
x
n
t = e) ((Xx
1
x
n
t) (e)) I2
2. (Xx
1
x
n
t = e) (Xx
1
x
n
t e) 1. {}/
3. .(Xx
1
x
n
t = e) .(Xx
1
x
n
t e) 2. NEC, K, MP
4. (Xx
1
x
n
t = e) .(Xx
1
x
n
t = e) NI
5. (Xx
1
x
n
t = e) .(Xx
1
x
n
t e) 3, 4. MP
6. !e((Xx
1
x
n
t = e) .(Xx
1
x
n
t e)) 5. UI
Corollary: !e(e = e) .(e e)
While necessary logical equivalence is not to be regarded as identity, identity does entail necessary
logical equivalence.
T28: .!eXx
1
x
n
t.(e Xx
1
x
n
t)
1. !x.(e e) .(e ro
1
o
n
) e1 .(e e)/
2. .(e ro
1
o
n
) Xx
1
x
n
t.(e Xx
1
x
n
t) EI
3. !x.(e e) PL, NEC, UG
4. Xx
1
x
n
t.(e Xx
1
x
n
t) 1-3. PL
5. .!eXx
1
x
n
t.(e Xx
1
x
n
t) 4. UG, NEC
T29: !X
n
x
1
x
n
!te.(e Xx
1
x
n
t)
1. (Xx
1
x
n
t) e(e/Xx
1
x
n
t) EI
2. .(Xx
1
x
n
t Xx
1
x
n
t) e.(e Xx
1
x
n
t) (1, .(ro
1
o
n
t Xx
1
x
n
t)/)
3. .(Xx
1
x
n
t Xx
1
x
n
t) PL, NEC
4. !X!x
1
x
n
te.(e Xx
1
x
n
t) 2, 3. MP, UG
T30: .(Xx
1
x
n
t e(e ^ .(e Xx
1
x
n
t)))
1. .(Xx
1
x
n
t Xx
1
x
n
t) PL, NEC
2. Xx
1
x
n
t (Xx
1
x
n
t ^ .( Xx
1
x
n
t Xx
1
x
n
t)) 1. PL
3. (Xx
1
x
n
t ^ .( Xx
1
x
n
t Xx
1
x
n
t)) e(e ^ .(e Xx
1
x
n
t)) EI
4. Xx
1
x
n
t e(e ^ .(e Xx
1
x
n
t)) 2, 3. PL
5. Xx
1
x
n
t e(e ^ .(e Xx
1
x
n
t)) 4 PL
6. .!X!x
1
x
n
(Xx
1
x
n
t e(e ^ .(e Xx
1
x
n
t))) 5. UG, NEC

93

T31: Necessity of e-Involvement: e .e
1. e ((e = Xx
1
x
n
t) ^ (( = X \ = x
1
x
n
) \ = t) Def
2. e ((e = Xx
1
x
n
t) ^ (( = X \ = x
1
x
n
) \ = t) 1. PL
3. ((e = Xx
1
x
n
t) ^ (( = X \ = x
1
x
n
) \ = t) (.(e = Xx
1
x
n
t) ^ .(( = X \ = x
1
x
n
)
\ = t) NI
4. ((e = Xx
1
x
n
t) ^ (( = X \ = x
1
x
n
) \ = t) e 1. PL
5. .((e = Xx
1
x
n
t) ^ .(( = X \ = x
1
x
n
) \ = t) e 4. K
6. e .e 2-3, 5. PL
T32: Analytic Entailment: !ee!e((e ee) (ee e))
(From ee5 by PL)
T33: Priority Anti-Reflexivity: (e < e)
By Def. Event Priority and t1.
T34: Priority Transitivity: ((e
1
< e
2
) ^ (e
2
< e
3
)) (e
1
< e
3
)

By Def. Event Priority and t2.
T35: ((e
1
< e
2
) \ (e
2
< e
1
)) \ (e
1
e
2
)
By Def. Event Simultaneity and t3.
T36: Eee-Comprehension: e
1
((e
1
) ^ e
1
t) ee!e
2
((e
2
ee) ((e
2
) ^ e
2
t))
This is an instance of the ee2 comprehension schema.
T37: Non-Vacuity of Time: .!te(e ^ et)
1. !x!tX Xxt T27
2. X(Xxt ^ (Xxt = Xxt)) 1. CQT, I1.
3. eX(e ^ (Xxt = e)) 2. T26
4. ((e = Xxt) ^ (( = X \ = x) \ = t)) e Def , PL
5. e(e ^ et) 3-4. PL
6. .!te(e ^ et) 5. UG, NEC
It is impossible for there to be a time at which no event occurs.
T38: Maximal t-Event: .!tee!e((e ee) (e ^ et))
1. e
1
(e
1
^ e
1
t) ee!e
2
((e
2
ee) (e
2
^ e
2
t)) ee2
2. !te
1
(e
1
^ e
1
t) !tee!e
2
((e
2
ee) (e
2
^ e
2
t)) 1. UG
-

94

3. !te(e ^ et) T38
4. !tee!e
2
((e
2
ee) (e
2
^ e
2
t)) 2-3. PL, NEC
At any world w and time t there is some compound event that includes every event involving t at w. Such
an event is effectively a stage of the world in question.
The next few theorems are basically direct implications of the ee comprehension schema.
T39: x-Stage: .!xee!e((e ee) (e ^ (et ^ ex)))
At any world w and time t, and object x, there is some compound event that includes every event
involving both t and x at w. Such an event is a complete description of x a given moment in time (in a
given world).
T40: Maximal x-Process: .!xee!e((e ee) (e ^ ex))
A maximal x-process is effectively a process to which every simple event involving x belongs. It provides
a complete description of x relative to the world in question.
T41: World Process: .ee!e((e ee) e)
A world process contains every occurent event within a world. Each world w can effectively be identified
with the world process that occurs in w.
The next two theorems are instances of the Pee schema.
T42: x-Segment: ee((ee ^ (pee ^ ee[x]) ee((Pee ^ (ee ee)) ^ !e(e ee (e
^ex))))
An x-segment is a process that contains every event involving an object x during a certain span of time.
T43: World Segment: ee((ee ^ pee) ee((Pee ^ (ee ee)) ^ !e(e ee e)))
The next three theorems are analogs of the reflexivity, antisymmetry, and transitivity of the subset
relation of set they.
T44: Reflexivity of Inclusion: !ee(ee ee)
1. ee = ee I1
2. (ee = ee) ((e ee) (e ee)) I2
3. (e ee) (e ee) 1-2. PL
4. (e ee) (e ee) 3. PL
5. !e((e ee) (e ee)) 4. UG
6. !ee(ee ee) 5. Def , UG

95

T45: Antisymmetry of Inclusion: ((ee
1
ee
2
) ^ (ee
2
ee
1
)) (ee
1
= ee
2
)
1. (ee
1
ee
2
) ((e ee
1
) (e ee
2
)) Def.
2. (ee
2
ee
1
) ((e ee
2
) (e ee
1
)) Def.
3. ((ee
1
ee
2
) ^ (ee
2
ee
1
)) (((e ee
1
) (e ee
2
)) ^ ((e ee
2
) (e ee
1
)))
1, 2. PL
4. (((e ee
1
) (e ee
2
)) ^ ((e ee
2
) (e ee
1
))) !e((e ee
1
) (e ee
2
))
PL,CQT
5. !e((e ee
1
) (e ee
2
)) (ee
1
= ee
2
) ee4
6. ((ee
1
ee
2
) ^ (ee
2
ee
1
)) (ee
1
= ee
2
) 3-5. PL
T46: Transitivity of Inclusion: ((ee
1
ee
2)
^ (ee
2
ee
2
)) (ee
1
ee
3
)
1. (ee
1
ee
2
) ((e ee
1
) (e ee
2
)) Def.
2. (ee
2
ee
3
) ((e ee
2
) (e ee
3
)) Def.
3. ((ee
1
ee
2
) ^ (ee
2
ee
3
)) (((e ee
1
) (e ee
2
)) ^ ((e ee
2
) (e ee
3
)) )
1, 2. PL
4. (((e ee
1
) (e ee
2
)) ^ ((e ee
2
) (e ee
3
)) ((e ee
1
) (e ee
3
)) PL
5. ((e ee
1
) (e ee
3
)) (ee
1
ee
3
) Def.
6. ((ee
1
ee
2
) ^ (ee
2
ee
3
)) (ee
1
ee
3
) 3-5. PL
Finally, with the introduction of events, the most precise formulation of DMC can be introduced:
A1: DMC: !e(e .e)
The logic of events herein presented is quite rich in both its syntax and semantics. This is, of
course, due to the fact that events possess a relatively fine-grained structure, namely the structure of
atomic temporal first-order formulas. Further explorations of its theorems must be saved for another time
(save for those which are of direct relevance the general theory of necessary substances). The logic is also
semantically very rich and it is both beyond the task of this treatise and my present abilities to rigorously
discuss its semantic properties. It is, of course, quite obvious that any axiomization will be incomplete
since it involves both second-order quantification over predicates and second-order quantification over
events themselves.



Copyright Adam Labecki 201

96

CHAPTER SIX:
FACTUAL EQUIVALENCE, GROUNDING,
AND QUALITATIVE PROPERTIES

Set theory enjoys the fact that, beyond some basic logic, it requires only two primitives: set and
membership (or element) one primitive kind and one primitive relation. Many philosophers are
envious of the mathematicians capacity for reductive analysis. Philosophers typically have a harder time
building all of their concepts upon the basis of one or two primitives and usually when they try, many of
the resulting definitions are subject to various counterexamples or charges of circularity. An admirable
attempt was made by Lewis (1983, 1986a) to define a cluster of central concepts by taking naturalness as
the sole metaphysical primitive. Natural is a higher-order property and natural properties are, roughly
speaking, those which make for objective similarity (and differences) between concrete objects.
Generally speaking, it is perhaps asking too much of a metaphysical system that it should be able
to define all of its distinctive terms by means of one or two primitives. Given the rather formalized nature
of the present work, however, the pressure to relay on a small number of primitive terms is especially
strong. So in this respect, I must be more demanding of myself than I am on my peers.
One of the primitives to be introduced will effectively be something like natural, in the sense of
natural property, namely qualitative, in the sense of qualitative property, will be introduced. I shall also
have recourse to two additional primitives: factual equivalence and grounding. Factual equivalence is
practically a logical relation, except for the fact that it presupposes that there are facts of a more or less
worldly sort. Grounding, as I shall use the term, ultimately comes to something like causation in
the sense of generation or production. Both factual equivalence and grounding can be construed as
operators and hence will receive a model theoretic treatment. Quality, being a non-defined second-order
monadic property, will not be directly featured in the semantics.
1. Grounding and Factual Equivalence
We often say that one thing is so because of, or in virtue of, or something else. Likewise, metaphysicians
often like to say that one thing or class is metaphysically (or ontologically) prior to, or more fundamental
than, some other thing or class. Recently, a number of metaphysicians, most notably Fine (2001) and
Schaffer (2009),
79
have proposed grounding as a, or even the, central primitive term in metaphysics.
Shaffer, for instance, deems grounding to be the primitive structuring conception of metaphysics

79
Also see: Rosen (2010), Bennett (2011), Schnieder (2011), and Audi (2012). For a critical, but friendly, discussion
of grounding, see Sider (2011 Ch. 8).

97

(Schaffer 2009) and defines a cluster of central concepts in terms of grounding. Likewise, Fine (2012b)
insists that questions regarding grounds are central to realist metaphysics.
Historically speaking, the first distinctive reference to something like metaphysical grounding
appears to consist in Aristotles claim that one thing can be prior by nature to some other. In the Politics
he famously argues that the polis is prior by nature to both households and individuals (1253
a
20). In
Metaphysics he offers four accounts of priority where both the first and the fourth accounts of prior
include priority by nature. The first is with respect to being closer to some beginning in an absolute
and natural way. This account is rather oblique and it is no doubt the sort of talk that raises suspension
regarding the notion of ground in the first place. What does it mean to be closer to some beginning in an
absolute and natural way? One way to understand this is to suppose that priority here consist in being
closer to those things which are fundamental and basic. If substances are the basic entities, as they
apparently are for Aristotle,
80
then substance is prior to accident and accident is posterior to substance.
Likewise, an infima specie, such as horse, is said to be more fundamental than its genus, mammal, since
it is closer to individual substance. The fourth kind of priority is priority in nature and substance
and is essentially a dependence relation, where the posterior cannot exist without the prior but the prior
can exist without the posterior. This, of course, is a modal notion of ground which has enjoyed favor in
recent decades, typically under the name supervenience, at least when supervenience is taken to be
asymmetrical. In recent years, however, many metaphysicians have come to agree with Fines (1996)
contention that modal dependence is too coarse grained to be of much explanatory value. Hence there has
been a recent attempt to rehabilitate something like Aristotles first definition of the natural priority.
81

Grounding is generally understood in terms of non-causal explanation (Audi 2012) and it is
typically illustrated by the following sort of examples:
The mental is grounded in the physical
Truths are grounded in states-of-affairs
Wholes are grounded in their parts
Dispositional properties are grounded in categorical properties
Moral truths are grounded in non-moral truths
In the interest of keeping primitives to a minimum, I might have preferred to speak of a general
explanation relation. Bennett (2011a) makes a similar proposal when speaking of building relations,

80
In the course of the Metaphysics, however, Aristotle seems to have revised his views on this matter, where
something like substantial essence or substantial form comes to be regarded as that which is most primary.
81
For more on priority in the philosophy of Aristotle see Peramatzis (2011).

98

where causation may be included as a kind of building relation. The problem, however, is that causal
explanations do not behave in the same way as non-causal explanations, and so each will require its own
axioms and semantics.
Bennett notes that a major difference between causation and many other building relations,
such as material composition, is that the former is diachronic while the latter is synchronic. This is
certainly true, at least given the commonly held assumption that causes are temporally prior to their
effects, but I think that the more crucial difference here is that an effect is surely something over and
above its cause but, in most cases of grounding, the grounded is regarded as nothing over and above its
ground. This, in turn, leads to a modal difference. When the grounded is nothing over and above its
grounds, it would seem to be determined and hence necessitated by its ground. In the case of causation,
the cause is productive of its effect but might well have obtained while producing a different effect.
I would also classify ontological, as opposed to epistemological, emergence, if there is such a
thing, as a form of causation. For instance, if there are emergent mental properties or minds, such entities
would be a genuine product of the interaction of the physical properties of the brains parts. This would
especially be so if the mind as such were emergent from the brain. I believe that part of the skepticism
toward emergence is the result of skepticism toward productive causation. Humeans find the idea that one
entity could actually generate another, distinct, entity positively mystifying. Indeed, they probably find it
mystifying that the collision of two high-momentum particles can result in more than two particles. But
this seems to happen and, as Heil (2012) notes, it would seem to be a kind of emergence.
Indeed, the primary problem with most discussions of grounding is that it is typically believed
that the grounded is nothing over and above its grounds (Fine 2001, Schaffer 2009, Rosen 2010) and
yet it is neither a part of its ground nor identical with it! As in the case of modal supervenience, it is
thought that once you have the grounds, the grounded items are an ontological free-lunch (Armstrong
1997). Fine (2001) suggests that doing realist metaphysics requires some basic grasp of what this is
supposed to mean. I must admit that, the more I reflect on the matter, the weaker becomes my grasp.
In this case, I say, there really is no such thing as a free lunch. If the Bs are really nothing over
and above the As, that just means that the Bs are just the As differently conceived, on a different level of
analysis, or simply described in different vocabulary, etc. For instance, if a whole is really nothing over
and above its parts, then the worldly (i.e., non-linguistic) facts about the whole and the worldly facts
about the parts constitute the same facts. In each case we may employ different vocabulary, but we are
just talking about the same facts in different ways.

99

Audi (2012) rejects the idea that the grounded is nothing over and above its grounds. In doing so
he denies that many putative cases of grounding are genuine, but are, rather, cases of factual equivalence:
Rosen rejects instead the worldly conception of facts. This seems to me to yield a very different
picture of grounding from the one that I am after, one on which what grounds what is determined
at least as much by our concepts as it is by the antecedent structure of the world. What we get
from grounds is, at least in cases of reduction, not something responsible for the grounded fact, but
rather a more accurate conceptual rendering of the very same worldly state-of-affairs (Audi 2012:
111).
It seems, then, that there are two distinct conceptions of ground. One is a matter of which worldly facts
about the world are more fundamental than others and one about which conceptualizations of identical
facts are more basic or fundamental. I am sympathetic with Audi on this matter in that I believe that the
central primitive concept of metaphysics should be about fundamental reality, not about how we talk
about things. I differ from Audi, however, in that I am uncertain whether or not there are any forms of
explanation which are both non-causal and non-reductive.
82

So I shall replace many putative grounding relations with factual equivalence which is, of course,
a symmetrical relation. If the whole is nothing above and beyond the sum of its parts, it is also the case
that the parts are nothing above and beyond the whole. But usually when one affirms the former, one
means to say that the parts are ontologically more basic or fundamental than the whole. But, again, I think
Audi is right to suggest that that the fundamentality in question here is not really ontological but
conceptual. For instance, I side with Merricks (2001) in affirming that a whole must be something above
and beyond the sum of its parts and if there are no such wholes then there are no wholes at all (i.e.,
mereological nihilism would be true).
2. Qualitative Properties
A qualitative property is a property that is intended to describe what an object is like. Talk of
qualitative properties, however, admittedly suffers from the fact that our conception of quality seems to
arise on the basis of something like qualia. The problem is that qualia describe what it is like to
experience this or that feature of an object. This leads us to fear that what it is like only really applies to
consciousness and when we are talking about objects being like this or like that, we are projecting mental
properties onto the things themselves. Qualia (supposing one accepts such things) are subjective in their
very nature. There are qualities that we attribute to objects based on which qualia they produce in our
mind; colors serve as the most common example. There is the qualia red (or rather a family thereof)
and, when something produces it in me, I say of that thing that it is red. But the thing is not really red, at

82
Audis putative examples of non-causal and non-reductive explanations mostly pertain to value theory rather than
metaphysics.

100

least if the meaning of red is closely associated with the qualia red. Red only describes how something
is relative to conscious beings with certain physiologies under certain kinds of background conditions. To
be sure, that something appears to be red is indicative of how something is disposed to reflect and absorb
light and, given that dispositions do not float about on their own, appearing red is indicative of what
something is like in itself. A red object primarily reflects a certain part of the light spectrum and mostly
absorbs other the rest of the spectrum (at least within the light-spectrum visible to humans). The way
something is such that it mostly only reflects red light is probably not an irreducibly qualitative feature
of the object. It probably has more to do with certain structural features of its surface.
Many Early Modern philosophers such as Descartes and Locke believed that the only genuinely
objective properties were those that can be described mathematically. More generally, we may speak of
structural properties which are descriptive of the way in which the parts of an object are arranged and
bonded together. In chemistry and biology the behavior of objects are typically explained in terms of such
structural properties. Structures alone only provide partial explanations, since the causal capacities of the
components provide the remaining explanation. As Searle (1980) famously pointed out, a system of water
tubes arranged in the manner of a human brain does not yield a mind. The structure needs to be made of
the right stuff. Moreover, if there are elementary particles, differences between them cannot be explained
in terms of structure at all. In such cases, we must either posit primitive dispositional difference, a
particularly brute form of primitive modality, or we must presume that there is some qualitative difference
between members of the two kinds, in virtue of which they have the powers that they do. I shall denote
qualitative properties by the second-order predicate .
Returning to the original concern, however, our notion of quality seems to arise entirely from
our familiarity with qualia. This seems to be where we get our conception of what the object is like
from. But since qualia are mental objects, we become uncertain whether or not we can even conceive of a
wholly non-mental object. This line of reasoning led Berkeley to his Idealist conclusions. Idealism might
be resisted if we resist certain empiricist assumptions about how we arrive at our ideas. That is, we might
try to deny that our idea of quality is completely determined by our experience of qualia; perhaps it is
also, in part, innate. Or perhaps we can abstract from qualia a general conception of quality that goes
beyond the phenomenal nature of qualia themselves.
Categorical is frequently contrasted with dispositional (to be discussed below). Roughly
speaking, categorical properties characterize the way something is at some given time, while
dispositional properties concern how something might be affected or affect some object under specified

101

conditions.
83
Martin (1998), Heil (2003), and Jacobs (2011) particularly identify the categorical with the
qualitative, while Ellis (2001) usually has structural properties in mind when he speaks of categorical
properties. I think that it is safe to say that all categorical properties are either qualitative or structural. It
will, however, be possible to define structural property in terms of partial factual equivalence, so the
only primitive needed here will be qualitative property in the sense of a pure quality that does not
depend upon any structure.
3: The Logic of Factual Equivalence and Ground
Fine (2012a) has made a major contribution to the fledging field of the logic and semantics of ground,
84

where grounds is treated as a binary operator. His treatment of ground bears some similarity to van
Fraassens (1969) treatment of facts and tautological entailment and Yablos (1982) study on semantics
for grounding and dependence. There are, however, some difficulties associated with Fines treatment.
First, his semantics cannot evaluate the truth values of iterated grounding formulas (i.e., it only permits a
first degree object language with respect to the grounding operator). This means we cannot say of a
grounding formula that it is grounded in something else. This is theoretically somewhat costly, since it
would be nice to isolate some fundamental base that grounds all non-fundamental truths whatsoever and it
does not seem that grounding relations can, in general, be perfectly fundamental.
85
Second, his semantics
do not interact with modality.
It should also be noted that Fine distinguishes four distinct grounding relations. In addition to the
distinction between full and partial ground, where a partial ground offers a partial explanation, he
introduces a distinction between strict and weak ground. We end up with the following sorts of ground:
strict full, weak full, strict partial, and weak partial. Weak grounding is conceived in terms of
explanations which permit reflexivity and symmetry. Hence we can always say that is weakly grounded
by and we can say that Ian is married to Jessica grounds the fact that Jessica is married to Ian. Such
cases, however, appear to be cases of factual equivalence. Accordingly, I shall replace weak grounding
with factual equivalence, which will, however, be axiomatically and semantically distinct from full weak
grounding. Correia (2010) makes use of a factual equivalence operator as part of his treatment of the logic
of ground (although he does not identify it with any species of what is commonly taken as ground).

83
The term categorical property was likely chosen due to the fact that dispositional properties might be thought of
as hypothetical or conditional properties, since dispositional predicates at least conceptually entail the way
something might be or act under typically counterfactual circumstances, while categorical properties characterize
how something is in a non-conditional, or categorical, manner.
84
Correia (2010) has offered a somewhat different approach.
85
See Sider (2011).

102

I have suggested that most common examples of metaphysical grounding are, in fact, matters of
factual equivalence. The facticity here is assumed to be worldly or substantial rather than linguistic or
conceptual. On this conception, Ians being married to Jessica and Jessicas being married to Ian
should count of factually equivalent, when true. These two propositions express the same fact or state of
affairs. With the introduction of time-index they refer to the same events. It will turn out that events that
are not strictly identical can turn out to be factually equivalent. That is, while e
1
and e
2
might pick out
distinct members of , it may be that said that said members are nevertheless made-true by the world in
the same way. This has turned out to be the case because events were introduced on the basis of atomic
temporal formulas or structured propositions. In this respect they turn out to be too fine-grained for some
purposes.
Only those facts that occur are capable of being factually equivalent. Non-occurent events can be
possibly factually equivalent, or conditionally so, but never are they so simpliciter. Hence, while the logic
of events and processes outlined in the previous chapter lends itself to a somewhat linguistic conception
of events, factual equivalence assists in treating those events which obtain as worldly events.
The following axioms for factual equivalence are largely based on those presented in Angell
(1977) and Correia (2010).
F1: ( )
F2: ( ( ^ ))
F3: ( ) ( )
F4: ( ^ ( ^ )) (( ^ ( ^ )) (( ^ ) ^ ))
F5: ( ^ ( ^ )) (( ^ ( ^ )) (( ^ ) ^ ))
F6: ( ) (( ^ ) ( ^ ))
F7: (( ) ^ ( )) ( )
F8: ( ) ( )
F9: ( ) ( ^ )
Full ground, for which I shall use the symbol , can be defined in terms of factual equivalence
and one-way necessitation.
is short for (( ) ^ (.( ) ^ .( ))
Given this definition, full groud will be assymetrical and hence also irreflexive. Hence most of
the axioms of factual equivalence will fail. Indeed, only F7-F10 will hold for full grounding.
T47: ( ) ( )
T48: ( )

103

T49: (( ) ^ ( )) ( )
T50: ( ) ( )
T51: ( ) ( ^ )
T52: (( ) ^ ( )) ( ( ^ ))
I shall use the symbol to denote the partial ground operator, where reads: is
partially grounded by or partly because of .
PG1: ( ) ( )
T53: ( )
PG2: (( ) ^ ( )) ( )
PG3: ( ) ( ^ )
PG4: (( ) ^ ( ))
Likewise, strong factual equivalence, for which I shall use the symbol , may be defined as
follows:
is short for (( ) ^ .( ))
T54: ( ) ( ( ^ ))
1. ( ) (( ^ ) ( ^ )) F6
2. ( ) F9
3. ( ( ^ )) F2
4. ( ) ( ( ^ )) 2-3. PL
5. ( ) ((( ^ ) ( ^ )) ^ ( ( ^ )) 1, 4. PL
6. ((( ^ ) ( ^ )) ^ ( ( ^ )) ( ( ^ )) F7
F10: ( ) (( ) ( ))
F11: ( ) (( ) ( ))
Semantics
Recall that in the semantics for events that the is the set of all events, actual, possible, and
impossible. Let V, be a support frame where V is a subset of () and is a total preorder on V (i.e.,
satisfies the properties of reflexivity, transitivity, and totality). is to be understood as a temporal
ordering of events and is effectively determined by the times included in the variable assignments of
events. The symbol will be used when is symmetric, whereby two members of V will be equivalent

104

with respect to their temporal ordering. For any world w and such that
w
, let [ ]
w
be the support
function assigning to a subset of V, denoted by []
w
subject to the following condition:

For any , if t then for every ee and t if [ee] e []
w
and t e [ee], then t t
Now let consider the difference between synchronic and diachronic support, beginning with the
following definition.
Maximal: Given a frame A, R, x is maximal in A under R just in case, for every y e A, yRx
Synchronic Support: For any the synchronic support of , denoted as []
wS, is the set of all
members of []
w
that are maximal under .
For any true in w, []
wS verifies . That is, each member of []
wS collectively make it the case that is
true.
Diachronic Support: For any , the diachronic support of , written []
wD, is the
complement of []
wS in []
w
.
For a given , each member [ee] in []
wD, enables (or potentiates) . Given [ee] e []
w
, the members of
[ee] collectively possess the power or potential to bring it about that the synchronic support should
eventually obtain. Intuitively speaking, each member of []
wD is a complete immediate or mediated
cause of []
wS. Assuming that causation is indeterminate, members of []
wD do not necessitate , since
in some possible worlds such events may occur without ultimately bringing it about that is the case.
This difference may also be explained by the fact that is something over and above any of its diachronic
supports.
The synchronic and diachronic support of a formula form a simple partition of []
w
.

The
synchronic support will effectively be the truth-maker of , where is nothing over and above any
member of []
wS.
In cases in which is a negation, say = (Fxt ^ Gxyt) the support function will be
understood to exclude the possibility of the truth of the formula of which it is a negation, in this case
(Fxt ^ Gxyt). Any synchronic support for such a formula would be one which effectively excludes
one or both of the conjuncts from being true. I am, in effect, assuming that sets of positive facts exclude
other facts as a matter of metaphysical necessity. For instance, if is interpreted as a has up-spin at t,
then is falsified, and hence is verified, if, as a matter of fact, a has down spin at t. Likewise, if is
interpreted as this rose is red, then is falsified if, as a matter of fact, the rose in question is yellow.
Whatever makes it the case that the this rose is yellow is true also makes it the case that this rose is

105

red is false (where both instances of this refer to the same rose). Likewise, the events in virtue of
which x has up spin at t are also the events in virtue of which it does not have down spin at t.
One might object that certain exclusions are contingent upon nomic laws or, worse, some
exclusions might be contingent to particular cases. My response to such objections is as follows. The
fundamental verifiers of a state of affairs might involve quite a few events, including those which
effectively ground contingent laws. Regarding nomological contingency, I do not believe in brutally
contingent laws (this shall be discussed in more detail in Chapter Eight). Effective laws might have
differed due to different universal background conditions, most notably differing values of the
cosmological constant, but they could not have just differed. The ultimate grounds, on the fundamental
level will include such global conditions, since they do contribute to the phenomena in question. If one
remains skeptical of the exclusion approach to negative facts, one might well include negative facts
within , and hence V, but it is very doubtful that there could be negative fundamental facts.
In the interest of tractability, it is important to show how the support functions of complex
formulas are related to the support functions of their components.
Let both atomic first-order formulas such as Fat and simple events be atomic formulas and let
be a metavariable for atomic formulas.
[
1
^n]
w
= {[ee] | {[
1
]
w
,,[
n
]}/ }
That is, [
1
^n]
w
is the class of all simultaneous members of the set consisting of the union of the
support set of each of its conjuncts. Compound events are verified in a similar fashion in terms of their
members.
[
1
\
n
]
w
= {[
i
^j]
w
| {
i
,,
j
} , {
1
,,
n
} and V
M
(
i
^
j
, w) = 1}
That is, each member of [
1
\
n
]
w
is the support set of the conjunction of any subset of its obtaining
disjuncts. The support sets of single disjuncts are included since is
i
is true, then so is
i
^
i
.
The support sets for negated disjunction and negated conjunctions follow a similar pattern:
[((
1
...\
n
)]
w
= [
1
^]
w

[(
1
^n)]
w
= {[(
i
\j )]
w
| {
i
,,
j
} , {
1
,,
n
} and V
M
((
i
\
j
)

, w) = 0}
[!x(x)]
w
= {{[(a
1
)]
w
,,[(a
n
)]
w
}/ }
(Where a
1
a
n
is a sequence of every object in D.)
86


86
Matters would be somewhat complicated if there the cardinality of D is uncountable.

106

That is, [!x(x)]
w
is the equivalence class of the union of the support set of each instance of something
satisfying , where it happens to be the case that everything does satisfy .
[x(x)]
w
= {[(a
i
)^(a
j
)]
w
| V
M
((a
i
)^(a
j
), w) = 1
That is, [x(x)]
w
is simply the set of the support set each of its conjunctive (including idempotent
conjunctions of single cases) instances. Given the above, the analysis of combining quantification and
negation is straightforward:
[!x(x)]
w
= {[((a
i
)\(a
j
))]
w
| V
M((a
i
)\(a
j
), w) = 1
[!x(x)]
w
= {{[(a
1
)]
w
,,[(a
n
)]
w
}/ }
Analyzing the support sets of modal formulas in terms of their component formulas is a bit more difficult.
The most straight forward approach is a follows:
[.]
w
= {{[]
w1
,,[(a
n
)]
wn
}/ }
(Where w
1
w
n
is a sequence of every w in W)
That is the fundamental facts that make it the case that is necessarily true is the temporal equivalence
class of the union of the fundamental facts that make the case at each world in which is true, being all
of them. Likewise for possibilities:
[]
w
= {{[]
wi
,,[]
wj
}/ | V
M
(, w
i
) = 1 and V
M
(, w
j
) = 1
This analysis of the facts that make possibilities true will only be satisfactory, however, within the
framework of modal realism. Facts obtaining in other possible worlds can only make things true if those
facts are not merely representational but genuine concrete states of affairs. Actualists would wish to
effectively restrict members of V to events that occur in the actual world.
My own suggestion is that the truthmakers of modal formulas should consist in the dispositions or
powers involved in actually occurring states of affairs. Here is one way a formula like .Fxt could be
made true. Suppose there is some first moment t at which x exists and, at t the categorical state of x is
such that x is F and in any way that the world might have evolved from t onward, x would continue to be
F. Likewise, .Fxt would be verified if xs initial disposition is such that x is not F and no way in
which the world might evolve after t could result in x becoming F. The matter is a bit tricky if there is no
time in which x begins to exist. Given permantism, this will be so for every object should there be no
absolute beginning of time. In such cases any member of [.Fxt]
w
might have to include facts about x
during every point at which it exits, during each which time x is F and has no potential to cease being x.

107

Making this notion of modal truth-making precise, however, would typically require the introduction of
branching world semantics (Chapter Eight). But sufficient precision may be obtained as follows:
Let us assume that V only includes events that occur in the actual world and let P(V) be a
function that assigns each member of V with a subset of (). Intuitively, for each ee such that [ee] e V,
P([ee]) is the set of events that could potentially be caused by the occurrence of ee, immediately or
transitively.
[]
w
= {[
1
^
n
]
w
| []
w e P([
1
^
n
]
wS)}
[.]
w
= {{[
1
]
w
,,[
n
]
w
}/ | {[]
w1
,,[]
wn
}

, P([
1
^
n
]
w
}
Now that the analysis of the support sets of complex formulas in terms of their components has
been completed, we may now proceed to make use of support sets in the truth-valuations of the primary
operators that are to be introduced.
Initial Segment: Set A is an initial section of set B under R if and only for some x e A, which
is non-maximal under R, B = {y e A | yRx}.
[V] V
M
( , w) = 1 iff []
w
= []
w

[V] V
M
( , w) = 1 iff there is a bijective function f from []
w to []
w
such that for each
[ee
1
] e []
w
, [ee
1
] f[ee
1
]
[V] V
M
(

, w) = 1 iff []
w is an initial segment of []
w

[V] V
M
( , w) = 1 iff there is a bijective function f from []
w to an initial segment of []
w

such that for each [ee
1
] e []
w
, [ee
1
] f[ee
1
]
The valuation for factual equivalence () simply assigns the same support set to each of the two
connected formulas. The valuation for partial grounding () requires that each member of the support set
of the partial ground should be a proper subset of some member of the support set of the partially
grounded item and, conversely, each member of the partially grounded item should contain every event
contained in some member of its partial ground. The valuation of for full generation () requires that the
support set of the generator should consist of an initial segment of the support set of the generated item.
Finally, the truth valuation for partial generation () requires that the support set of partial generator
should effectively be a partial ground of a full generator. Full and partial generation will effectivey be
employed to denote full and partial causation between events.
4. Some Work for Factual Equivalence, Grounding, and Qualitative
Df47: Intrinsic: For any property X, X is intrinsic, X, if and only if:

108

!x!t(Xxt !Y
n
!y
1
y
n
((Xxt Y
n
y
1
y
n
t) (Py
1
xt...^...Py
n
xt))
87

X is an intrinsic property if and only if, for any x, if Xxt is partly grounded by the obtaining of some
relation Y
n
then each (object) relatum of Y
n
y
1
y
n
t must be a part of x.
Df48: Extrinsic: For any property X, X is extrinsic, written X if and only if:
!x!t(Xxt Y
n
y
1
y
n
((Xxt Y
n
y
1
y
n
t) ^ (Py
1
xt...\...Py
n
xt))))
88

X is an extrinsic property just in case, for any x, if Xx then Xx is partly explained by the obtaining
of some relation Y
n
, at least one relatum of which is mereologically distinct from x. By these definitions,
every property is either intrinsic or extrinsic.
Df49: Internal Relation: a relation X
n
is internal, written X
n
, if and only if:
!x
1
!x
n
!t(X
n
x
1
x
n
t Y
1
Y
n
((Y
1
^ Y
n
) ^

(X
n
x
1
x
n
t (Y
1
x
1
t^Y
n
x
n
t))))
An internal relation is a relation that is factually equivalent to some intrinsic states of its relata.
Df50: External Relation: a relation X
n
is external, written, X
n
, if and only if it is not
internal.
Df51: Structural Property: A property X is structural, written X, if and only if:
!x!t(Xxt y
1
y
n
Y(((Y ^ (Py
1
xt^Py
n
xt)) ^ (Xxt Yy
1
y
n
t)))
Df52: Categorical Property: A property X is categorical, written X, if and only if
X \ X
Event Causation:
Full and partial event causation will be denoted in terms of full and partial generation
respectively.

Kim (1974) discusses various ways in which one event might depend on another event but
not caused by it. This occurs in simultaneous events, i.e. as when a rake is moving some leaves while an
arm is moving the rake. A second way in which an event might depend on another yet not be caused by
another is if one is a part of another.

87
Rosen (2010) provides the following two-part definition of Intrinsic, using grounding as a primitive. For any
property X, X is intrinsic, X, if and only if:
a) .!x((Xx (y)) Pyx)
b) .!x(Xx (y)) (y)
The second clause is added in order to ensure that lonely, the property of being the only concrete object, cannot
turn out intrinsic (which is an already a doubtful possibility, given NNE).
88
The second conjunct is similar to the definition of extrinsic proposed Francescottis (1999), who calls it a d-
Relational property. Francescotti uses the phrase consist in in his definition, which is indicative of grounding, but
he ultimately takes consist in to be a relation of identity so as to ensure explanatory relevance.

109

GN1: Asymmetry of Causation: (ee
1
ee
1
)

GN2: Irreflexivity of Causation: (ee
1
ee
2
) (ee
2


ee
1
)
GN3: Transitivity of Causation: (ee
1


ee
2
) ^ (ee
2
ee
3
) (ee
1
ee
3
)
The same also, of course, holds for partial causation. The following axiom governs the interation between
full and partial causation:
GN4: Causal Subsumption: ((ee
1
ee
2
) ^ (ee
3 ee
1
)) (ee
3
ee
2
)
T54 (ee
1
ee
2
) (ee
1
^ ee
2
)
Df53: Immediate Cause: Eee
1
is the immediate cause of Eee
2
, written ee
1
I

ee
2
if and
only if:
(ee
1
ee
2
) ^ !ee
3
((ee
3


ee
2
)

((ee
3


ee
1
) \ (ee
3
= ee
1
)))
Df54: Immediate Partial Cause: ee
1
is an immediate partial cause of ee
2
, written ee
1

I
ee
2
if
and only if:
(ee
1

ee
2
) ^ ee
3
((ee
3
I

ee
2
) ^ (ee
1
ee
3
))
Df55: Causal Process: A plurality of events, ee
1
is a causal process, written P
C
ee
1
, if and
only if:
Pee
1
^ !ee
2
!ee
3
(((ee
2

t
ee
1
) ^ (ee
3

t
ee
1
)) ^ (ee
2
ee
3
)) ((ee
2
ee
3
) \ (ee
3
ee
2
))
A process Pee is a causal process just in case any two distinct stages belonging to Pee are such that one
must be at least a partial cause of the other. Generally speaking, I will only be interested in causal
processes and so I will usually just say process when speaking of a casual process.
Df56: Closed Causal Process: A plurality of events, ee
1
is a closed causal process, written
P
IC
ee
1
, if and only if:
P
C
ee
1
^ !ee
2
!ee
3
(((ee
2

t
ee
1
) ^ ((ee
2

t
ee
1
)) ((ee
2
ee
3
) \ (ee
3
ee
2
))
Metaphysically speaking I believe that total causation is more fundamental than partial causation, but we
tend to be pragmatically more interested in partial causation than total causation. Knowledge of partial
causation permits us to know what we need to add or remove from a larger situation in order to produce
some desired result. In our practical reasoning we are typically not interested in total causes. Moreover,
there is also the difficulty of discerning when a causal process is causally isolated from the rest of the
universe, at least with respect to the production of a specific effect. Indeed, if the world is sufficiently

110

holistic every independent causal process will be a segment of the entire world. That is, it could be that
total immediate causation only obtains between successive world-stages.
Do21: Universal Causation 1 (UC1): !ee
1
ee
2
(ee
2
ee
1
)
Do22: Universal Causation 2 (UC2): !ee
1
ee
2
(ee
1
ee
1
)
UC1 is the doctrine that every event has a total cause. In most respects UC1 is a non-controversial
doctrine, but since causes are temporally prior to their effects, UC1 entails that there is no first moment of
time. Indeed, some might regard this as a good reason to permit simultaneous causation. Theists are likely
to be sympathetic with the spirit of UC1 but might insist that we permit some sort of agent causation
whereby the first temporal moment is caused not by an event, since events are occurents, but by God or,
more specifically, the divine will in conjunction with the divine power. While I do not believe that agent
causation is incoherent, I do believe that it is of little assistance in making sense of libertarian freedom. If
the effect is not produced by some event involving some agent but merely by the agent as such, then it
would seem to be grounded simply on the very existence of the agent. In the case of divine creation, it
would not permit us to understand how the act of creation is a free act, nor would it permit us to
understand how the temporal universe would be generated with a first moment of time rather than no first
moment. At any rate, if time does have a beginning one might want to designate a t
0
that would be
minimal in T but interpreted as representing the timeless state that gives rise to the whole, properly
temporal, series.
Likewise UC2 entails that time cannot come to an end, since each stage of the universe is bound
to generate a further stage. Eschatological religious doctrines aside, UC2 would appear to follow if some
from the conservation of energy and if conservation should be metaphysically necessary then UC2 would
surely be true.
I am inclined to believe that both UC1 and UC2 are true but I do not see how they could be
derived by more fundamental assumptions and their only interesting consequences would appear to be
those just discussed and so I would not like to introduce them as axioms.
5. Dispositions
Dispositional predicates such as fragile conceptually entail some sort of conditional or other. If
this vase is fragile is true, then it follows, of conceptual necessity, that there are circumstances under
which this vase would shatter. Now suppose that vases are fragile in virtue of possessing a certain
monocular structure with property F. Having monocular structure F does not conceptually entail that

111

something is disposed to shatter when dropped. Hence, it is commonly argued, dispositional predicates
and their bases cannot be identical.
This is certainly true. But predicates are one thing and genuine properties are quite another
(Armstrong 1978, Lewis 1983). I follow Shoemaker (1980) in distinguishing between dispositional
predicates, on the one hand, and dispositional properties, on the other. For instance, being made of
copper, while not a dispositional predicate, does, at least as a nomological fact, bestow capacities upon
its bearer, such as the capacity to conduct electricity. Hence being made of copper is a powerful
property it bestows certain capacities upon its bearer. Shoemaker (1980), and others who believe that
genuine properties are powerful properties, suggests that we do not restrict dispositions or powers to
dispositional predicates but regard as dispositional any property that comes with some set of powers.
Hence, when I speak of dispositional properties I do not mean dispositional predicates so much as
powerful properties I do not mean fragile so much as the way vase is such that it is fragile.
There are two things to note here. First, the dispositional property is itself some way that an
object is and, second, the way is particularized. That is to say, not every fragile thing is fragile in virtue
of the same property, although most such properties are likely to be similar.
Dispositional ascriptions, which are entailed by dispositional predicates and supported by
dispositional properties, are typically understood relative to a form of conditional analysis. For instance,
when we say that a vase is fragile, we mean that there is some condition c under which it would break,
where the breaking is the manifestation of the vases fragility. Both circumstances and manifestations
shall be treated as fine-grained, or determinate, events. For instance, shattering as such is not fine
grained there are countlessly many ways in which objects may shatter and one and the same object will
shatter in different ways under different conditions. If a manifestation is described in terms shattering a
perfectly determinate shattering is indented. Roughly speaking, dispositional ascriptions are functions
from circumstances to manifestations. It should be noted that, in addition to being wholly determinate, c
and m would typically be compound events.
Dispositional Ascription: A dispositional ascription, written D
(cm)
x, picks out exactly one
determinate circumstance, c, and manifestation, m, entailed by the possession of some
dispositional property.
The most basic analysis of dispositions goes by the name Conditional Analysis (CA):
Conditional Analysis: x is disposed to m under conditions c, written D
(cm)
x just in case x would
m were it the case that c.

112

CA runs into certain problems such as finks (Martin 1994) and masks (Bird 1998). A disposition is
finkish if it is possible that an occurrence of c coincide with or is quickly followed by occurrences by
which the disposition in question is removed before it has had the chance to manifest itself. A more
sophisticated, and somewhat cumbersome, form of CA, usually denoted SCA has been proposed by
Lewis (1997):
Something x is disposed at time t to give response r to stimulus s iff, for some intrinsic property B
that x has at t, for some time t after t, if x were to undergo stimulus s at time t and retain property
B until t, s and xs having of B would jointly be an x-complete cause of xs giving response r
(Lewis 1997: 157).
89

An x-complete cause is a cause that is complete relative to the intrinsic properties of x. In addition to
addressing the question of finks, SCA also stipulates that the disposition in question must be grounded in
some intrinsic property of x. It is typically thought that the grounding property in question should not
itself be dispositional, but categorical.
While SCA apparently addresses the problem of finks, it is not evident that it can address the
problem of masking. A disposition is maskable if it is possible for c to obtain while m does not obtain
even if the disposition in question is not removed. Masks are sometimes called antidotes since the most
common example of masking is the case of a poison being masked by the presence of an antidote. Lewis
attempts to hold onto SCA by making the circumstances more specific so as to preclude the presence of
any masks. Indeed, one might have already responded to finks in this way, in which case CA was fine all
along. The concern here, however, is that we have started along a road toward evoking ceteris paribus
clauses that will ultimately render CA and SCA vacuous. That is, one begins to fear that CA will
ultimately amount to something trivial like: x is disposed to m when under the conditions c, unless it is
not. Bird (2007) and Hauska (2009) suggests that we embrace building (mostly implicit) ceteris paribus
clauses into CA. In as sense, my own proposal to will involve getting specific, but it will not be in the
form of CA or SCA.
Fara (2005) has suggested the following analysis of dispositions in terms of habituals rather than
conditionals:
HABITUAL: x is disposed to m when c just in case, should x m when c, it would be in virtue of
some intrinsic property of x that it would do so.
HABITUAL enjoys the fact that it matters not if the disposition is finkish, maskable, or, for that matter,
probabilistic (which will turn out to be quite important). It simply tells us that if x should m when c, then
it is in virtue of some intrinsic property of x, rather than for some extrinsic reason. The in virtue of

89
Lewis, of course, writes s where we write c and r where we write m.

113

relation involved in Farras definition, of course, should be read as a grounding relation and, in particular,
partial causation, since the circumstance c typically involves some dispositional partner or other. My own
analysis, however, shall differ from Ferras in that the property in question need not be intrinsic, although
it will need to be perfectly determinate and I shall later add that it must be natural.
An additional virtue of habitual analysis is that it is capable of handling probabilistic dispositions
or propensities. Indeed, given the probabilistic nature of present physics, it is likely that propensities are
indeed the norm. Some propensities are rather distinct from other kinds of dispositions in that they do not
require a circumstance distinct from an object being disposed in a certain way over a period of time. For
instance, a nucleus disposed to manifest beta decay does not do so in response to some change in its
circumstances. It is disposed to manifest beta decay, on average, once every so many units of time. One
might think that such dispositions are counterexamples to the kinds of analysis we have been considering.
On habitual, however, such is not the case. In such cases c is simply x having some property at some time
span t
m
t
n
along with any ambient condition that might also contribute to m (where such conditions
most likely also contribute to the holding of the dispositional property in question).
The fact that my definition of event does not require that the property instantiation should be
new is useful in that it permits analysis of non-triggered dispositions. If events necessarily involve
change, and if the causal relata are necessarily events, then one is forced to speak of indeterministic decay
events as if they were simply uncaused. I think that this is misleading and likely to make the
phenomena appear far more mysterious than is necessary. It is in virtue of the powerful properties of the
Sun a moment ago that it continues to exist and radiate light. The two events are by no means fully
distinct entities. Objects typically sustain themselves and most of the power had by the Sun a moment
ago continues to be present in it now. This should actually come as no surprise but, unfortunately, it
conflicts with the sacred Humean intuition that any two events are completely distinct and hence there can
be no necessary connection between any two events.
The following definition is a variation of habitual that I would like to propose:
Df57: Dispositional Analysis: An object x is disposed at t to m when c, written D
(cm)
xt, if
and only if:
X((X ^ Xxt) ^ ((c ^ Xxt)
I
m))
In words: an object x is disposed to manifest m under conditions c at t if and only if, x possesses some
wholly determinate property at t and, in some possible world w, the conjunctive event of c and Xxt
is the total immediate cause of m. c will typically be compound event and may include ambient conditions
(i.e. conditions that are typically present and normally taken for granted).

114

On the usual analysis of dispositions, a dispositional property grounds dispositional ascriptions
regarding how something would manifest itself upon the completion of a single counterfactual event.
Dispositions, however, also ground how something would manifest itself upon the completion of
sequence of events or a process. For instance, water, in its liquid form, has the disposition to boil when
heated but it does not have an immediate disposition to shatter. If, however, water is first frozen, it
acquires the disposition to shatter. A child has the capacity to learn French, but may not have actually
learned it. Knowing French is typically taken to be a dispositional property, but the ability to learn
French is also a disposition. So knowing French is something like an iterated or second-order
disposition that can be ascribed to someone who has not learned French but has the capacity to learn
French.
90
What I have been calling dispositions are ordinary non-iterated, or first-order, dispositions,
where something is such that a single event may trigger its manifestation. Roughly speaking, x has the
iterated disposition to manifest m just in case there exists at least one string of possible circumstances, at
the end of which, x may manifest m partly in virtue of having some determinate property X. Each c
i
in the
series marks an event in which the disposition of x might be altered and where the penultimate
circumstance provides x with the immediate or first-order disposition to manifest m.
For convenience, I will quantify over dispositional ascriptions, where D
(mc)
(D
(mc)
) is short-
hand for Xee
1
ee
2
(X
(ee1ee2)
).
Df58: Iterated Dispositional Analysis: An object x is disposed, at time t
1
, to m upon the
completion of a causal process cc, written D
(ccm)
x, if and only if:
X((X ^ Xxt
1
) ^ (Xxt
1
cc)) ^ (!t
2
(cct
2
t
1
t
2
) ^ (cc
I
m))
An iterated dispositional ascription is true of x at t just in case, X possesses some wholly
determinate property at t, the possessing of which is contained within the earliest stage of cc and it is
possible for cc to be a total cause of m, in which case Xxt would be a non-immediate partial cause of
m. The idea is that the possession of X by x at the beginning of the process is causally relevant to the
manifestation.
From a metaphysical point of view, iterated dispositional ascriptions are nothing exotic. The fact
that we are typically interested in non-iterated dispositional ascriptions is no doubt due to pragmatic and
conventional reasons.

90
Molnar (2003) briefly brings up the matter of iterated or second-order dispositions. See also: Ellis (2001),
Borghini and Williams (2008), and Mumford and Anjum (2011).

115

For the purpose of good book keeping, it is worthwhile to show how dispositional predicates and
properties are ultimately tracked by our semantics. It will be necessary to first treat dispositional
ascriptions and then treat dispositional properties.
The interpretation function for a dispositional ascription I (r
(cm)
) will assign both a standard
intension of a sub-set of D at each world w and a member of a set . Each member of is an ordered
triple ee
1
, ee
2,
e where [e] e while [ee
1
], [ee
2
], and are each subsets of , subject to the following
conditions:
1. [ee
1
] e
E
and [ee
2
] e
E

2. [e] e [ee
1
]
3. [ee
1
]
w
, [ee
2
]
w
and it is not the case that [ee
1
]
w
= [ee
2
]
w

4. For any t, t, if ee
1
t and ee
2
t, then t < t.
5. For any ee
3
if [ee
3
]
w
, [ee
2
]
w
and it is not the case that [ee
3
]
w
= [ee
2
]
w
, then [ee
1
]
w
, [ee
3
]
w
The first condition stipulates that ee
1
and ee
2
are compound events. The second that e is included in ee
1
.
The third and fourth that ee
1
is a total cause of ee
2.
The fifth that ee
1
is an immediate total cause of ee
2.

Relative to the dispositional analysis schema, ee
1
= c, ee
2
= m, and e = Xxt
1
.
The semantical analysis of iterated dispositions is similar except the condition is not an event
but a thick process. The interpretation function for an iterated dispositional predicate I(r
(ccm)
) will assign
both a standard intension of a sub-set of D at each world w and a member of a set
n
. Each member of
n

is an ordered triple ee
1
, ee
2,
e where, again, ee
1
is a thick process rather than a compound event.
Finally, for each dispositional property r, I (r) will assign both a standard intension of a sub-set
of D at each world w and a set consisting of members of and
n
. In this way, dispositional ascriptions
are like plural properties, each one being a plurality of dispositional predicates, standard and iterated. We
can then say that Xxt Dcm)
xt is true just in case D
(cm) X is true, likewise for iterated
dispositional ascriptions. This would also justify the axiom that two dispositional properties are identical
just in case they both contain the same dispositional predicates.
Df59: Dispositional Equivalence: Xx
t
is factually equivalent to a dispositional
ascription D
(cm)
x
t
, written D
(cm)
x
t
Xx
t
if and only if:
(X ^ Xxt) ^ ((c ^ Xxt)
I
m)
Obviously, an atomic formula can also be factually equivalent with an iterated dispositional ascription in
the expected manner.

116

Df60: Disposition: A property X is a dispositional property, written X, if and only if:!x!t(Xx
D
(cm)1
D
(cm)
(((D
(cm)1
xt Xxt)^(D
(cm)
xt Xxt)) ^ ((D
(cm)1.
x^D
(cm)n.
x) Xxt))) and
D
(ccm)1
D
(ccm)
!x!t(Xxt (((D
(ccm)1
xt Xxt)^(D
(ccm)n
xt Xxt)) ^ ((D
(ccm)1.
x^D
(ccm)n.
x)
Xxt)))
The possession of a dispositional property is factually equivalent to satisfying a number of dispositional
ascriptions, both individually and conjunctively and hence is functionally equivalent to a set of
dispositional ascriptions. In short, we can establish that a property is dispositional just in case its being
possessed is factually equivalent with the possession of a dispositional ascription:
Df61: Disposition (Simplified): X D
(cm)
!x!t(Xxt (D
(cm)
xt Xxt))
T55: !x!t!D
(cm)
(D
(cm)
xt X(X ^ (D
(cm)
x Xxt)))
1. D
(cm)
xt X((X ^ Xxt) ^ ((c ^ Xxt)
I
m)) Df. 61
2. (D
(cm)
x
t
Xx
t
) (X ^ Xxt) ^ ((c ^ Xxt)
I
m) Df. 62
3. X(D
(cm)
x
t
Xx
t
) X((X ^ Xxt) ^ ((c ^ Xxt)
I
m)) 2. CQT
4. D
(cm)
xt X(D
(cm)
x
t
Xx
t
) 1, 3. PL
5. X D
(cm)
!x!t(Xx (D
(cm)
xt Xxt)) Def. X
6. D
(cm)
!x!t(Xx (D
(cm)
xt Xxt)) X 5. PL
7. X(D
(cm)
x
t
Xx
t
) X(Xx (D
(cm)
xt Xxt)) PL, CQT
8. (Xx (D
(cm)
xt Xxt)) D
(cm)
!x!t(Xx (D
(cm)
xt Xxt)) CQT
9. D
(cm)
xt D
(cm)
!x!t(Xx (D
(cm)
xt Xxt)) 4, 7-8. PL
10. D
(cm)
xt X 5, 9. PL
11. D
(cm)
xt (X ^ (D
(cm)
x
t
Xx
t
)) 4, 10. PL
12. (X ^ (D
(cm)
x
t
Xx
t
)) X(X ^ (D
(cm)
x
t
Xx
t
)) CQT
13. D
(cm)
xt X(X ^ (D
(cm)
x
t
Xx
t
)) 11-12 PL
14. !x!t!D
(cm)
(D
(cm)
xt X(X ^ (D
(cm)
xt Xxt))) 13. UG
For every dispositional ascription there is some dispositional property such that instantiating said
ascription is factually equivalent to instantiating said property.
Before proceeding, it should be noted that the above analysis is not intended to provide an
analysis of dispositional predicates. For instance, an object that shatters under some, but very few,
conditions, would not usually count as fragile. One would have to say that something is fragile just in
case there is an appropriate number of the right sort of circumstances under which it would shatter.
Specifications here are most likely to be both vague and tedious. This simply provides additional reason

117

to suppose that ordinary dispositional predicates like fragile do not cut nature at the joints. Moreover,
when an object is disposed to shatter under certain circumstances, the property that grounds the shattering
ascriptions will most likely ground other sorts of manifestations under different circumstances. Perhaps
the way in which a vase is such that it would shatter under certain circumstances also enables it to contain
water the way that it does, or reflect light in the way that it does.



















Copyright Adam Labecki 2013




118

CHAPTER SEVEN:
SUBSTANCE AND POWER

The guiding thought behind the principles to be introduced in the following two chapters is that the world
is fundamentally a world of powerful particulars and by concrete object or substance I mean nothing
more or less than a powerful particular. This guiding thought does not constitute a picture since no
determinate kind of entity has been presupposed nor is it even assumed that fundamental reality is such
that it could be accurately represented by anything like a picture. Moreover, my use of the plural in
powerful particular should not even be taken to mean that I am assuming that there is a proper plurality
of such particulars, although I would surely be surprised if there were only a single substance.
1. The Eleatic Principle and Thick Realism (A2)
I suggest that anything has real being that is so constituted as to possess any sort of power either to
affect anything else or to be affected, in however small a degree, by the most insignificant agent,
though it be only once. I am proposing as a mark to distinguish real things that they are nothing but
power (Soph. 247d-e).
The doctrine outlined in the passage has been dubbed the the Eleatic principle, named after the speaker
who puts it forward in Platos Sophist. Several contemporary philosophers, including Armstrong (1978,
1997), Ellis (1990), and Heil (2003), have suggested that we should adopt some sort of Eleatic Principle
when doing metaphysics. In the broadest strokes, those who adopt the Eleatic Principle claim that causal
power is the mark of real being.
Different philosophers have drawn different morals from the Eleatic Principle. Armstrong, for
instance, applies the principle to deny that there can be any abstract objects, including transcendent, or
ante res, universals. For Heil, the moral is that all natural or sparse properties (i.e., all real properties)
are dispositional. The following discussion will be guided by the Eleatic Principle, interpreted in this
latter manner.
91
On my own account, power is the mark of concrete or substantial being and the non-
concrete must be grounded in the concrete (i.e., the non-powerful should be grounded in the powerful). So
rather than say that absolutely everything must be powerful, I shall claim that everything is either
powerful or depends upon the powerful.
In the final sentence of the passage cited above, however, the Eleatic Stranger seems to be suggesting that
one should be a bare dispositionalist.
Df62: Bare Disposition: A dispositional property X is said to be bare if and only if:

91
Not that such is taken to be an interpretation of the Platonic text at issue. On that score, Armstrong can be seen as
closer to the truth, since the Eleatic Stranger shows how the Principle will require that the friends of the Forms
revise their theory of Forms (i.e., transcendent universals).

119

Y(Y ^ !x(Xx (Xx Yx))
A bare disposition is a disposition the having of which is not factually equivalent to the having of some
categorical property. Such a disposition denotes a potential or ability that something just has, and there is
nothing more to say about the matter.
Do23: Bare Dispositionalism: X((X ^ Xxt) ^ Y(Y ^ (Xxt Yxt)))
Do24: Pan-Bare Dispositionalism: is the doctrine that all dispositional properties are bare.
Bare dispositions would make things possible for their bearers without, however, making any difference
to the way their bearers actually are. This would involve a strong form of primitive modality that many
philosophers would find unattractive. Nevertheless, many of those metaphysicians who take dispositions
seriously believe that there are bare dispositions
92
and a few, such as Bird (2007) believe that all
dispositional properties are bare. I believe that bare dispositions should be avoided if at all possible.
A2: !x!t.(X Y(Y ^ (Yxt Xxt)))
If X is a dispositional property then there exists some categorical property Y such, any object x that
possesses Y at t also possesses X at t.
To some readers A2 might appear to be too strong. Many philosophers believe that might
categorical properties contingently bestow the dispositional properties that they bestow. Again, I am in
disagreement with many philosophers on this matter. In my judgment categorical properties necessarily
bestow those dispositional properties that they do bestow and when we imagine contrary cases we are
either imagining worlds in which said dispositions are universally masked or we are imagining worlds
that are not actually possible. Again, I shall return to these matters in Chapter Eight. The primary reason
for holding A2, however, consists in being a realist regarding powers while holding that dispositional
properties cannot be bare. That is, A1 is intended to articulate the intuition that objects really are powerful
in virtue of the way they are.
T56: !X!x(X Y(Y ^ .!t(Yxt Xxt)))
1. .(X Y(Y ^ (Yxt Xxt))) A2, CQT
2. X Y(Y ^ (Yxt Xxt)) 1. T
3. (Xxt Yxt) .!t(Yxt Xxt) Def. , CQT
4. Y(Y ^ (Yxt Xxt)) Y(Y ^ .!t(Yxt Xxt)) 3. PL, CQT
5. X Y(Y ^ .!t(Yxt Xxt)) 2, 4. PL
6. !X!x(X Y(Y ^ .!t(Yxt Xxt))) 5. CQT

92
For instance Ellis (2001), McKitrick (2003), Molnar (2003), and Mumford (2006).

120

A2 is the first metaphysically substantive axiom to be introduced, the others pertaining to formal
languages except A1, which is the mere denial of modal collapse (although many readers will no doubt
regard F8 and F9 as substantial claims). I call this axiom Thick Realism in that it presupposes a thick
world of powerful particulars which are able to sustain themselves throughout time and generate effects
when interacting with other particulars.
93
The alternative is Thin Realism, according to which causation
is either a mere matter of phenomenal regularities (neo-Humeanism) or is dependent upon contingent
laws that somehow govern causation from the realm of second-order relations (nomic necessity). Many
of my reasons for rejecting Humeanism are given in Appendix A (although I will say a bit more below).
As for nomic necessity, it is the result of wanting to have ones cake and eat it too. The nomic
necessitist wants to say that laws are higher-order universals that actually enforce regularities such as
!x(Fx Gx). But if some higher order universal between F and G makes it the case that all Fs are
also Gs, why does it do so in some worlds but not others? Because in some worlds F and G are related by
while in others they are not? So the higher-order relation F and G is evidently not an intrinsic relation,
but some external relation. But how do universals themselves enter into external relations in the first
place? From a semantic point of view, the intensions of higher-order predicates is determined by the
intensions of first-order predicates, this suggests that first-order predicates are fundamental and that
higher-order predicates are derivative and determined by truths about the first order predicates. Indeed,
this places some suspension upon the whole idea of higher-order universals (as if suspension was not
already present) in the first place. The first-order universals alone would explain the second-order
predicates in question. But then we have effectively fallen back into the regularity theory of laws the
laws are determined by the distribution of first-order properties among objects.
Accordingly, I believe that nomic necessity is not a viable option. Either physical laws are
effectively necessary (when the categorical circumstances in question are wholly specified) or they are
wholly contingent. That is to say, either propertied particulars really do make things happen (all by
themselves without the help of laws or Gods intervention) or they do not.
Something like Thick Realism has been endorsed by Martin (1997), Heil (2003, 2012), Strawson
(2008), and Jacobs (2011). These thinkers generally take dispositional properties and categorical
properties to be identical, although I say only that all dispositional properties are categorical, since the
properties of abstract objects seem to be categorical yet non-dispositional.
94


93
A conceptually different version of Thick Realism might have been deflationary regarding dispositional properties
as such but maintain that when a categorical property grounds a dispositional ascription, it does so necessarily.
94
I do not believe that there is actually any disagreement here in that the above mentioned authors concern
themselves only with natural properties when discussing these matters.

121

I shall try to keep the motivation of A2 to a minimum. Nevertheless, given the prevalence of neo-
Humanism among contemporary metaphysicians, one could easily compose a monograph motivating and
defending A2, and so I shall have to offer more by way of motivation that I would like.
2. Quiddities Lead to Radical Skepticism and Faade Realism
Separating the dispositional and the categorical leads to a fairly strong form of skepticism, since it
could be the case that, in the actual world, some distinct categorical properties are dispositionally
identical while some very similar categorical properties are dispositionally very dissimilar. Categorical
properties so understood are sometimes referred to as mere quiddities, by analogy with mere haecceities.
95

A mere haecceity is essentially a bare particular an individual without any interesting essential
properties. Now, let R be a variable to denote roles. Roles are rather rich in necessary properties. Now, for
bare particulars, it is clearly the case that for objects x and y, and world w if Rx, w and Ry, w are true,
then there exists some world w such that Ry, w and Rx, w are true. That is to say, objects can
swap roles between possible worlds and, indeed, every role is played by every object at some world or
other. Many regard this as a good regarded as a good reason to reject haecceities or bare particulars
altogether.
96

When the categorical and dispositional are disconnected, a similar implication follows for
categorical properties, only the roles in question are the dispositional properties conferred by
categorical properties. In Lewis (2009) this conclusion is based upon his principle of recombination.
97
If
intrinsic duplicates of anything can exist or fail to exist with intrinsic duplicates of anything else, then the
laws of nature need to be contingent, and hence the causal roles of properties had better vary between
possible worlds. In particular, Lewis has it that there can be complete swapping in the dispositional roles
played by properties. Many commentators regard quiddities to be just as ontologically unsavory as
haecceities.
98

The most basic motivation for Thick Realism might be the conviction that similarity and
difference between objects can be discerned by their behavior, at least when objects are observed in
highly controlled environments. If (categorical) properties and powers are independent of one another, it
is not clear that we have any reason to infer categorical similarity from dispositional similarity. If the

95
See R. Black (2000).
96
Most famously, Chisholm (1967).
97
Haecceities do not arise from recombination on Lewis MR, since he makes use of counterpart theory as an
alternative to transworld identity, but quiddities do arise, since he understands transworld identity to obtain in the
case of properties.
98
See Shoemaker (1980, 1998), R. Black (2000), and Bird (2007).

122

categorical and the dispositional are completely independent of one another, I have no justification in
believing that two cats are more similar to one another than either is to the Moon. One cannot be any kind
of scientific realist if one believes that the categorical and the dispositional are completely independent of
one another.
Let us call this line of objection to quidditism quiddistic skepticism. Schaffer (2005) attempts
to defend quidditism from this line of attack. He suggests that the neo-Humean should respond to such
scenarios in the same way in which one typically respond to other skeptical scenarios: The solution to
the brain-in-a-vat and dreaming scenarios is not to decree them metaphysically impossible (!), but rather
to explain how knowledge can withstand them (Schaffer 2005: 19). In particular, he claims that
quiddistic skepticism is just another species of external world skepticism and one should respond to it in
the same sort of way in which one responds to other forms of skepticism regarding the external world.
It seems that there is an important disanalogy between quiddistic skepticism and typical forms of
external world skepticism. Take the brain in a vat scenario, for example. Let us suppose that such
scenarios are indeed metaphysically possible. That is, there are some worlds in which agents experience
themselves as if they lived in a material world in much the same way that we do, but are in fact brains in
vats. What I want to suggest is that the possible worlds in which such scenario obtain are rather sparse
compared to those in which they do not. That is to say, we rightfully call them remote possibilities, not
because their truth would be so epistemically devastating, but because they are metaphysically unlikely.
Indeed, most possible worlds in which there are brains in vats are probably also worlds in which only a
minority of brains are in vats as opposed to bodies.
99
That is, it seems that such scenario involve rather
artificial conditions which are parasitic upon natural conditions which do not give rise to skeptical
scenarios. Brains in vats are not natural phenomena, they are (at least typically) only found in worlds in
which brains are typically within organic bodies.
Let be a subset of including only possible events (i.e., events that occur in at least one
possible world and let A be the subset of containing all possible events in which an agent strongly
believes himself to be physically interacting with external objects. Let S be the subset of A in which one
would count as belonging to a skeptical scenario. The claim is that members of S are significantly more
rare (or infrequent) than members of the compliment of S in A (although it is possible that both would be
of the same infinite cardinality). Let us call this the Modal Rarity response to radical forms of skepticism.
Like all replies grounded in modal intuitions, it is certainly fallible. It is based upon the assumption that
some states of affairs are intrinsically more natural than others. One thing to note about the Modal Rarity

99
This could include mechanical bodies provided that the brain in question is receiving informative sensory inputs
from its immediate environment.

123

response is that, if we were in such a scenario, our intuitions regarding natural possibilities might be
radically misguided as well. This might be so if we happen to be in some sort of scenario in which
Subjective Idealism or Occasionalism are true and turn out to be metaphysically necessary.
Now let us turn to quiddistic skepticism. Is it plausible that these scenarios are rare in comparison
with their epistemically indiscernible non-skeptical counterparts? Reflection upon the nature of quiddities
quickly leads one to conclude that such scenarios, far from being rare, are the norm should it be the case
that quidditism be true.
If quidditism is true, so far as we know, there could be a dozen categorically distinct natural kinds
within the class of entities we have named electron but it so happens that each of these classes
contingently play one and the same dispositional role in the actual world. Indeed, each of these classes
might have no categorical properties in common. This would mean that our term electron does not cut
at the categorical joints of nature. Again, we cannot infer categorical joints from behavior. It gets worse.
Suppose that it is also the case that several classes of objects fall under the term photon. Now let, E be
one subclass of electron-behaving entities and let P be one subclass of photon-behaving entities, it could
be that E and P form a partition of some categorically natural class C. That is, some photons and some
electrons might actually be categorically identical! After all, if there is no necessary connection between
the categorical and the dispositional, why should every member of C contingently behave the same way
in the actual world? Recombination itself will guarantee this. If there is a world with a C behaving
electron-like and a world with a C behaving photon-like then there is a world that includes duplicates of
both states of affairs. There may be some worlds in which every C-object contingently plays the same
dispositional role, but this would be a rare exception. We can go further. There is no reason why the
dispositional role played by a categorical property should not alter from moment to moment. That is, it is
possible for something to switch from behaving like an electron to behaving like a photon, not due to any
change in itself or even its environment, but it might just happen to switch roles for no reason at all.
We can go still further. It could be the case that there is but one categorical property, different
instances of which play different causal roles. Everything is categorical the same as everything else, but it
is contingently the case that different instances play different dispositional roles at different times.
Alternatively, it could be the case that no two objects share any natural property at all. Every object is
completely different from every other object, but some happen to play similar dispositional roles.
If this sounds just wrong then so should the existence of quiddities. Hence, if quidditism is true,
then possible scenarios of quiddistic skepticism are the overwhelming norm. Cases in which they obtain
are highly frequent and cases in which they do not are abnormal even among those worlds that seem to
exhibit causal regularity. That is, quidditism does not simply admit the possibility of a kind of skeptical

124

scenario, it makes the belief that there is any correspondence between dispositional being and categorical
being no more justified than the belief that a given set of numbers will be drawn in the next State lottery.
That is, if we admit the possibility of quiddistic skepticism, we should also assume that we are in such a
scenario, but the same cannot be said for other sorts of skeptical scenarios, but the same cannot be said for
other kinds of skeptical scenarios.
Once the connection between the categorical and the dispositional is made wholly contingent, we
quickly find that we have no cognitive grip on the subject of categorical properties. The assumption that
we can track categorical similarity by tracking dispositional similarity is really all that we have to go on,
at least as far as empirical knowledge of the world is concerned. In the absence of such correspondence
we have no hope of grasping even a silhouette of objective reality.
100

Regarding Humes famous denial of necessary connections between distinct entities, A2 denies
the assumption that categorical properties are ontologically distinct from dispositional properties. Again,
dispositional ascriptions are merely predicates, and in ordinary language they tend to be rather imprecise
ones at that. A dispositional property, on the other hand, is the way that something is such that it satisfies
a set of wholly determinate dispositional ascriptions.
Thin Realism also appears to be unjustified as a form of realism about the external world. If I do
not believe that my perceiving of any object x is partly grounded upon the causal powers of x, what does
it mean to believe that x is a real concrete material object rather than a merely phenomenal being? Given
Humeanism, none of my perceptions are really caused by external objects acting on my sense organs and
the state of my sense organs does not give rise to the states of my brain. Hence, the existence of an
external world, at least one that extends beyond my brain, does not explain why I have the sorts of
perceptions that I have. But the basic case against phenomenalism is that perceptions and mind are not the
sort of things that can just exist by themselves. Phenomena are supposed to be at least partially
grounded in external objects. But in this case grounding really does seem to require causation. If
external physical objects are not partially causing the phenomena then they are not explaining anything at
all and hence fall prey to Occams Razor.
If one wants to block A2 by denying powers altogether, then one ought to be a phenomenalist if
not an outright subjective idealist.
101
Oddly enough, most neo-Humeans only follow their master when it

100
The idea is that, even if we cannot know intrinsic categorical properties in themselves, then we can still track
differences in the categorical being between object in tracking their dispositional being. We may, by controlled and
repeated observation, roughly know how many categorical joints there are, which ones tend to be found together,
which ones do not tend to go together, etc. the resulting being something like grasping the silhouette of the world
in itself.
101
See Molnar (2003): 123.

125

comes to powers and causality. They seem to think that one can be a Humean regarding causation while
still being justified believing in the existence of an objective physical world. If we are going to be radical
empiricists lets just go all the way and refrain from positing anything beyond perceptions.
Separating the categorical and the dispositional as in Lewis (2009) only encourages skepticism
regarding the very existence of categorical properties. Hawthrone (2001), regarding quiddities, asks
Why posit from the Armchair distinctions that are never needed by science? (2001: 369). Indeed, as
Armstrong (1968) and Ellis (2001) observe, it does appear that all of the properties of fundamental
physics are irreducible dispositional. At any rate, what we know of their dispositionality is exhaustive of
what we know about them. Hence, if the categorical and dispositional are not merely conceptually
distinct, but are also ontologically distinct, then it will be difficult to motivate the need for categorical
properties at all.
3. Some Work for Thick Realism
3.1 The Metaphysics of Dispositions and Natural Properties
T57: No Bare Dispositions: X(X ^ Y(Y ^ (Yxt Xxt)))
1. X Y(Y ^ (Yxt Xxt)) A2. T
2. X \ Y(Y ^ (Yxt Xxt)) 1. PL
3. (X ^ Y(Y ^ (Yxt Xxt)) 2. PL
4. !X(X ^ Y(Y ^ (Yxt Xxt)) 3. CQT
5. X(X ^ Y(Y ^ (Yxt Xxt))) 4. CQT
T58: !D
(cm)
!x(D
(cm)
x Y(Y ^ (D
(cm)
x Yx))
1. D
(cm)
xt X(X ^ (D
(cm)
x Xxt)) T55
2. X(X ^ (D
(cm)
xt Xxt)) X(X ^ Xxt) Fact , CQT
` 3. !X(X Y(Y ^ (Yxt Xxt))) A2
4. X X XY(Y ^ (Yxt Xxt))) 3. CQT
5. D
(cm)
xt X X 1. CQT, PL
6. D
(cm)
xt XY(Y ^ (Yxt Xxt)) 4-5. PL
7. !D
(cm)
!x(D
(cm)
x Y(Y ^ (D
(cm)
x Yx)) 6. CQT
Df63: Natural Property: For every X, X is natural, written X, if and only if:
X ^ X

126

It has become common, since Armstrong (1979) and Lewis (1983, 1986a), to make a fundamental
distinction between abundant and sparse properties (and relations), where sparse properties are typically
referred to as natural properties: Any two things share infinitely many properties, and fail to share
infinitely many others (Lewis 1983a: 13).
102
The general distinction, of course, is ancient, going back to
Plato (Phaedrus 265d-266a) but it had gone out of style for some time, having resurfaced in the works of
some of the new-old metaphysicians.
Abundant properties do not capture the notion of resemblance, nor are they relevant to the causal
powers of things. Hence, in order to do metaphysics, we need to distinguish an elite group of natural
properties that carve nature at the joints. The following two passages are representative of Lewis view on
natural properties:
Because properties are so abundant, they are undiscriminating. Any two things share infinitely
many properties, and fail to share infinitely many others. That is so whether the two things are
perfect duplicates or utterly dissimilar. Thus properties do nothing to capture facts of resemblance.
Likewise, properties do nothing to capture the causal powers of things. Almost all properties are
causally irrelevant, and there is nothing to make the relevant ones stand out from the crowd.
Properties carve reality at the jointsand everywhere else as well. If it's distinctions we want, too
much structure is no better than none. (Lewis 1983: 346)
Sharing of [sparse or natural properties] makes for qualitative similarity, they carve at the joints,
they are intrinsic, they are highly specific, the sets of their members are ipso facto not entirely
miscellaneous, there are only just enough of them to characterize things completely and without
redundancy (Lewis 1986a: 60).
Qualitative already hits on the notion of naturalness but it is likely to comprise a subset of natural
properties, assuming that there will be natural structural properties. Moreover, quality does not yet tell us
anything about causation. By the time we reached dispositional we already reached a class of properties
that has turned out to be co-extensive with natural properties, but we have axiomatically stipulated that all
such properties must also be categorical. Categorical qualified by powerful seems to get us natural.
Do26: Pan-Dispositionalism is the doctrine that all natural properties are dispositional
properties.
Pan-dispositionalism is involved in the very definition of natural property provided above, which
definition is motivated by A2.
T59: Dispositional Essentialism: !X! D
(cm)
((X ^ (D
(cm)
x Xx)) .(Xx D
(cm)
x))
1. (D
(cm)
x Xx) .(Xx D
(cm)
x) Def

102
Page numbers for Lewis (1983) are from its reprinting in Lewis (1999). In claiming that just about any two
objects will have infinitely many properties in common and fail to share infinitely many properties, Lewis is
assuming the cardinality of the domain of objects is no less than
2
.

127

2. (X ^ (D
(cm)
x Xx) .(Xx D
(cm)
x)) 1. PL
3. !X!Y(X ^ (D
(cm)
x Xx) .(Xx D
(cm)
x)) 2. PL, CQT
Every natural property necessarily has the dispositional profile that it in fact has.
By Dispositional Essentialism, also known as Scientific Essentialism, it is established that every
natural property grounds the same dispositional ascriptions upon its bearers in every possible world.
The usual objection to Depositional Essentialism is that it does not seem to be compatible with the
widespread intuition that the laws of nature are contingent. I have already shown how the usual way of
making laws contingent comes at a great cost, namely quidditism. In Chapter Eight, I will directly
address intuitions regarding the contingency of the laws and nature. For the present, I only suggest that
any world that differs from the actual world with respect to its apparent laws of nature is one in which
certain global events or processes obtain which, as a matter of fact, do not obtain in the actual world. In
other worlds, the claim is that the fundamental laws are fixed across all possible worlds and what many
metaphysicians take to be fundamental laws (and fundamental objects) are actually derivative and are
contingent precisely due to their derivative nature.
Df64: Concrete Object: For every x, x is concrete, at time t, written Cxt, if and only if:
X(X ^ Xxt)
An object is concrete just in case it has at least one natural property concrete objects are precisely those
objects which are capable of doing things and can do things in virtue of the way they are.
Df65: Abstract Object: For every object x, x is abstract, written Ax, if and only if it is not
concrete, i.e., if and only if it does not possess any natural properties.
3.2 On Concreta and Abstracta
Most often concrete is defined in terms of being spatiotemporal, but I do not think that defining
it in that way is very helpful. It suffers from the difficulty that it might not include subatomic particles,
which frequently lack a determinate position in spacetime
103
and it would also preclude the existence of
non-physical concreta Cartesian minds, angels, or God. While contemporary metaphysicians frequently
presume that there are no such things, ones basic ontological categories should at least include a place for
such things, were they to exist, especially if such entities are regarded as metaphysically possible.

103
Resnik (1990) considers various ways in which existence in spacetime might be construed and finds each of
them wanting even in the case of physical objects.

128

I have instead opted to define concrete in terms of natural properties, where such properties are
both powerful and categorical. Abstract objects are those that possess categorical properties but no
dispositional properties they are wholly actual. Despite the intuitiveness of this claim, one might
question whether or not abstract objects are totally lacking in dispositional properties. Rosen raises this
difficulty:
Suppose John is thinking about the Pythagorean Theorem and you ask him to say whats on his
mind. His response is an event: the utterance of a sentence; and one of its causes is the event of
John thinking about the theorem (Rosen 2012).
Perhaps the notion of ground will come in handy here. We might ask if the Pythagorean Theorem has any
intrinsic properties the possession of which partially grounded Johns utterance. I am not sure if our initial
intuitions will be very certain here. I shall try to explain why we should not think that any intrinsic
property of the Pythagorean Theorem grounds the event which is Johns response.
On the Representational Theory of Mind (RTM), we can deny that the Pythagorean Theorem
itself is causally involved in in Johns response while affirming that a token of the Pythagorean Theorem,
a mental concreta, is involved.
104
Now if the mental concreta is partially caused by the Theorem itself we
will still be in trouble due to the transitivity of causation. That is, we have pushed the event of Johns
verbal act back to his thinking act. If his thinking is an event partially caused by some other event, that
event must involve the Theorem. But what event could that possibly be? Unless it is some event whereby
the Theorem itself is acting upon Johns mind, but in that case we are speaking about abstract objects as if
there were concreta and no one believes that abstracta do such things.
If one does not want to commit to RTM or mental concreta, one might adopt some kind of
adverbial account of intensional states, according to which John replied as he did because he was
thinking Pythagorean-Theorem-wise. On this view of cognition, representations are not like individual
objects in ones mind as much as modes or states of the mind. Johns response was merely (partially)
caused by his state of mind, which state happens to resemble the Pythagorean Theorem in some way.
105

What caused the state of mind? Probably some prior state of mind in conjunction, perhaps, with certain
environmental factors. In either case, we may deny that the Pythagorean Theorem itself is involved in
Johns response.

104
For instance, Fodor (1998) writes: RTM claims that mental representations are related to propositional attitudes
as follows: for each event that consists of a creatures having a propositional attitude with content P (each such
event as Jones believing at time t that P) there is a corresponding event that consists of the creatures being related,
in a characteristic way, to a token mental representation that has the content P (Fodor, 1998: 8). A similar replay
could be given in the case of computing. A computer program is an abstract object, but, if so, then no computer runs
the computer program itself but only some token thereof.
105
For a recent defense of adverbialism, see Kriegel (2008).

129

3.3 New Definitions: Completion of the Core Language
Now that natural has finally been provided with a precise definition, a number of new definitions come
into view.
Df66: Compositional Property: A property X is compositional, written X, if and only
if:
X ^ X
Compositional properties are natural structural properties. That is, they are natural properties that an
object has partly in virtue of its mereological structure.
Df67: Relational Property: A property X is a relational property, written X, if and only if:
X \ X
There are two ways in which a property may be relational. It may be extrinsic, in which case the property
is had partly in virtue of some relation its possessor bears with one or more distinct objects or it can be
compositional, in which case it is possessed in virtue of the parts of an object and relations that obtain
between them.
Df68: Atomic Object: An object x is atomic at a time t, written Sxt, if and only if
Cxt ^ !y(Pyxt y =
t
x)
Atomic objects are those objects which do not have any proper parts (i.e., those objects which are
mereologically simple).
Df69: Strongly Intrinsic Property: A property X is strongly intrinsic, written
s
X ,if and only if
X ^ X
Strongly intrinsic properties are those natural properties which are not possessed in virtue of the parts of
its bearer.
Df70: Fundamental Property: A property X is fundamental, written X, if and only if:
X ^
s
X
Fundamental properties might also be referred to as perfectly natural properties. The idea is that
the class of fundamental properties provides something like a minimal-base for the obtaining of other
properties.
The properties of the parts of an object are typically understood to be ontologically prior to the
properties to which they give rise and so a composite property can only possess a fundamental property X

130

if it does not possess X in virtue of any of its parts, although it is not entirely clear how such properties
would be possible.
The denial of fundamental compositional properties is suggestive of micro-physicalism and
surely this is part of why it seemed intuitive to Lewis. It does not, however, entail micro-physicalism.
First, it has not been established that the fundamental objects are even physical. Second, it could be that
physical systems might give rise to simple non-physical objects, such as emergent minds. Third, it could
be that there is only one concrete object, the universe itself.
Df71: Imperfectly Natural: A property X is imperfectly natural if and only if:
X ^ X
Df72: Fundamental Object: An object x is fundamental at time t, written xt if and only if:
X(X ^ Xxt)
Fundamental objects are precisely those objects which instantiate fundamental properties.
Do27 Priority Monism: x!t(xt ^ !y(yt y =
t
x))
Priority Monism is the doctrine that there is exactly one fundamental object.
Df73: Derivative Object: An object x is derivative if and only if it is not fundamental.
Df74: Derivative Concrete Object: An object is a derivative concrete object if and only if it is
derivative and concrete (i.e. if it possesses no fundamental properties but at least one relational
natural property).
Df75: Extrinsic Natural Properties: A property X is both extrinsic and natural if and only if:
X ^ X
Those of a Humean persuasion will likely believe that the class of extrinsic natural properties is
empty. For the neo-Humean extrinsic brings to mind properties such as having the same birthday as
John F. Kennedy or being taller than most residents of Bulgaria. At most, extrinsic properties might
result from holding external but natural relations with other objects, such as being 25 kilometers north of
the Thames, but such a property would not count as a natural property. Neo-Humeans holds this view
since it goes along with recombination. If perfect duplicates of x can exist or fail to exist along with
perfect duplicates of anything else, and if we imagine that said duplicates would appear just like the
originals (or near enough), then we must believe that things are just the way we perceive them to be in
virtue of themselves and themselves alone. One imagines, for instance, that for each object x there exists
some lonely world at which a duplicate counterpart of x is the only concrete object. Accordingly, there

131

would be a world in which the Eiffel Tower (or rather a duplicate thereof) is the only concrete object,
which object has the exact same shape, size, etc. as the Eiffel Tower of this world.
At any rate, as noted in Appendix I, non-locality in physics suggests that the set of extrinsic
natural properties is not empty. Moreover, it will later be shown that no observable property (at least none
that we have discovered) is intrinsic. Mass, for instance, is extrinsic since it depends upon interaction
upon the Higgs field. But this does not mean that mass must be construed as unnatural, it just means that
natural properties can be extrinsic.
Df76: Natural Relation: A relation X
n
(where n < 1) is natural, written X
n
, if and only if:
Yxt(((Y ^ Y) ^ Yxt) ^ y
1
y
n
(Yxt X
n
y
1
yt) ^ (Py
1
x...\...Py
n
x))))
or
Yxt((Y ^ Yxt) ^ y
1
y
n
((Py
1
x...^Py
n
x) ^ (Yx X
n
y
1
y
n
))
(It should be understood that, in either case, that some of the ys must be distinct from some of the
other ys.)
A natural relation is a relation, the instantiation of which is possibly partially factually equivalent
with the possession of an extrinsic natural property or possibly partially factually equivalent with the
possession of a natural compositional property.
I am tempted to also require that natural relations should be external. Standard examples of
natural relations, most notably spatiotemporal relations, are certainly external. They do not simply
supervene upon the monadic states of their relata. Entanglement relations would also appear to be
external. As for causal relations, when these obtain it would seem that they do so in conjunction with
either spatiotemporal relations or entanglement relations. That is, if x and y are such that a change in x
immediately brings about a change in y, then we should conclude that x and y are entangled.
Nevertheless, entanglement does not appear to be external in the same way as spatiotemporal
relations. After all, if there is such a thing as spacetime of some sort of another, spatiotemporal relations
between objects x and y would supervene upon the intrinsic state of any connected sum of spacetime
points in which both x and y are located. Nevertheless, when several particles are entangled, we might
attribute their being entangled to some property that is possessed by the physical system comprising of
just those particles (although it is difficult to understand how such a property could be a categorical
feature of the system in question). That is, any such attempt would probably require a strong form of
wavefunction realism.

132

What would an internal natural relation look like, if there was such a thing? It would effectively
amount to something like sympathy. For instance, we are aware of the phenomena of sympathetic
resonance and sympathetic oscillation. If one holds down the damper pedal of a piano (or a harp) and hits
middle C with some force, then the F below middle C will mildly resonate along that portion of itself
which produces the same pitch as middle C.
Such sympathy is party grounded on the internal states of the strings but it also depends upon the
spatial the proximity of the strings, since the sympathy depends upon a signal travelling through the
intervening atmosphere. Strings which are too far from one other or which are placed in a relative vacuum
will not resonate with one another. Internal natural relations would be like sympathetic resonance but
without the dependence upon an external and local medium of communication. That is, if x and y were
internally connected there would be something about the qualitative state of x and y such that certain
changes in x would result in corresponding changes in y (or vice versa). Just as external natural relations
permit objects to become externally connected in ways that enable causal interaction, internal natural
relations would permit objects to become internally connected in ways that enable causal interaction.
Indeed, entanglement bears some similarity to sympathy, and so it might, after all, be better classified as
an internal relation than an external relation.
While it might be the case that all natural relations are external relations, I shall not build this
into the definition of natural relation, since the notion of an internal natural relation seems to be
coherent and, indeed, quite fascinating.
T60: Natural Extrinsic Properties are Partially Grounded by Natural Relations:
((X ^ X) ^ Xxt) Y
n
y
1
y
n-1
(Y
n
^ (Xxt Y
n
y
1
y
n-1
xt))
1. !Y(Xxt(((X ^ X) ^ Xxt) ^ y
1
y
n
(Xxt Y
n
y
1
yt) ^ (Py
1
x...\...Py
n
x))))
Y
n
) Def. Y
n

2. Xxt(((X ^ X) ^ Xxt) ^ Yy
1
y
n
(Xxt Y
n
y
1
yt) ^ (Py
1
x...\...Py
n
x)))) Y
n

1. CQT
3. ((X ^ X) ^ Xxt) Y
n
y
1
y
n
((Xxt Y
n
y
1
y
n
) ^ (Py
1
x...\...Py
n
x))))
Def. X
4. ((X ^ X) ^ Xxt) (((X ^ X) ^ Xxt) ^ Y
n
y
1
y
n
(Xxt Y
n
y
1
y
n
))
3 PL
5. ((X ^ X) ^ Xxt) Xxt((X ^ X) ^ Xxt) ^ Y
n
y
1
y
n
(Xxt Y
n
y
1
y
n
))
4, CQT, PL, T
6. ((X ^ X) ^ Xxt) Y
n
2, 5 PL

133

7. ((X ^ X) ^ Xxt) (Y
n
^ Y
n
y
1
y
n
((Xxt Y
n
y
1
y
n
)) 3, 6 PL
8. ((X ^ X) ^ Xxt) Y
n
y
1
y
n-1
(Y
n
^ (Xxt Y
n
y
1
y
n-1
xt)) 7 CQT
Every instantiation of a natural extrinsic property is partially explained in virtue of the instantiation of
some natural relation.
T61: Compositional Property Instances are Partially Grounded by Natural Relations
(X ^ Xxt) Y
n
y
1
y
n
((Y
n
^ (Py
1
x...^Py
n
x) ^ (Xx Y
n
y
1
y
n
))
1. X (Xxt y
1
y
n
Y
n
((Y
n
^ (Py
1
x^Py
n
x)) ^ (Xx Yy
1
y
n
)))Def X
2. Xxt((X ^ Xxt) ^ Y
n
y
1
y
n
(Py
1
x...^Py
n
x) ^ (Xx Y
n
y
1
y
n
)) Y
n

Def. Y
n

3. (X ^ Xxt) y
1
y
n
Y
n
((Py
1
x^Py
n
x) ^ (Xx Yy
1
y
n
)) 1. PL
4. (X ^ Xxt) (X ^ X) ^ Xxt) Def. X
5. (X ^ Xxt) y
1
y
n
Y
n
((Py
1
x^Py
n
x) ^ (Xx Yy
1
y
n
)) 3-4 PL
6. (X ^ Xxt) ((X ^ Xxt) ^ y
1
y
n
Y
n
((Py
1
x^Py
n
x) ^ (Xx Yy
1
y
n
)))
5. PL
7. (X ^ Xxt) Xxt((X ^ Xxt) ^ y
1
y
n
Y
n
((Py
1
x^Py
n
x) ^ (Xx
Yy
1
y
n
))) 6, CQT, T
8. (X ^ Xxt) Y
n
2, 7 PL
9. (X ^ Xxt) (Y
n
^ y
1
y
n
Y
n
((Py
1
x^Py
n
x) ^ (Xx Yy
1
y
n
)))
5, 8 PL
10. (X ^ Xxt) y
1
y
n
Y
n
((Y
n
Py
1
x^Py
n
x) ^ (Xx Yy
1
y
n
)) 9, PL
Every instantiation of a natural compositional property is partially explained in virtue of the instantiation
of some natural relation among its parts. In this case one would certainly expect the relation in question to
be spatiotemporal.
Df77: Fundamental Relation: A relation X
n
is fundamental, written X
n
, if and only if:
X
n
^ !x
1
!x
n
(X
n
x
1
x
n
x
1
^ x
n
)
That is, a relation is fundament just in case it is a natural relation whose relata are all fundamental objects.
Df78: Natural Simple Event: A simple event e is natural, written e, if any only if
X(X
n
^ eX) (Where n > 0)
Just as the properties that interest us the most are natural properties, the events that interest us the
most are natural events. The best known problems that generally plague Kims account of event

134

disappear when considering natural events. For instance, there is not one natural event for Brutus
stabbing Caesar and one for Brutus stabbing Caesar violently. Both of these are just different ways of
describing one and the same natural event.
While it is easy to understand what makes a simple event natural, it is more difficult to discern
what makes a compound event natural. Containing only natural simple events would not seem to be
sufficient, since said events might be spatially scattered and causally unrelated. I think that the best way
to guarantee that a compound event Eee is natural is to require that each e
1
contained in ee is either (1)
immediately partially caused by some simple natural event e
2
or (2) an immediate partial cause of some
simple natural event e
2
.
Df79: Natural Compound Event: A compound event Eee is natural, written Eee, if and only
if:
(1) !e
1
(((e
1
ee) e) ^ e
2
And at least one of the following two criteria is satisfied:
(2) e
1
!e
2
((e
2
ee) (e
2

I
e
1
))
(3) e
1
!e
2
((e
2
ee) (e
1

I
e
2
))
So natural compound events are those which are composed only of natural events and all of whose
components either contritube to the occurance of some other event or are all partially caused to occur by
some single event. The second criteria is intended to guarantee that all of the components of a natural
event are connected in some natural manner and they are all so connected either in all contributing to
some happening or in all being the partial result of some one occurance.
Df80: Fundamental Simple Event: A simple event e is fundamental, written e, if any only if:
X(X
n
^ eX
n
) (Where n > 0)
A fundamental compound event will be just like a natural compound event except each contained
simple event is not only natural but fundamental.
Causal processes are already effectively natural enough to count as natural. A fundamental
process would be a causal process that only involves fundamental events.
Do28: Supersubstantialism is the doctrine that something is a fundamental object if and only if
it is a spacetime point.
!x(x x is a spacetime point)

135

T72: Supersubstantialism is false.
Inasmuch as spacetime points exist at all, and I doubt that they do, they cannot be objects, i.e., are not
members of D, since, by PE, each member of D exists eternally.
Df81: Fundamental Base: a plurality of events constitutes the fundamental base, written ee, if
and only if
!e((e ee) e))
4. Foundations (A3)
4.1 Statement and Immediate Implications
I shall now introduce the second major metaphysical principle. It bears significant similarity to
the following claim made by Lewis (1994):
To begin, we may be certain a priori that any contingent truth whatsoever is made true, somehow,
by the pattern of instantiation of fundamental properties and relations by particular things (Lewis
1994: 473).
Let us call this doctrine supervenient base, in that it requires that all contingent truths are determined by
some sort of fundamental base consisting of basic objects, properties, and relations. Lewis takes the truth
of this doctrine to be an a priori matter. The question is not whether we should accept supervenient base
but how we might specify it. Lewis, of course, goes onto endorse to specify the doctrine in terms in the
form of Humean supervenience, which places certain restrictions on what sort of natural properties and
relations there might be while introducing a micro-reductive physicalist ontology.
The principle I am about to introduce is, in some respects, much like supervenient base. There
are, however, three differences. First, supervenes is replaced with is factually equivalent to. Second, I
do not presuppose that the fundamental properties and objects are physical and, in this sense, it remains
unspecific. Third, my principle makes a stronger claim in that it is not limited to contingent truths, but
extends to any truth whatsoever (provided that it does not contain any grounding or factual equivalence
operators). I call this doctrine foundations.
A3: Foundations: .( ee((ee ^ ee) ^ ( ee)))
(Where cannot contain any grounding or factual equivalence operators)
A3 is intended to express the view that all truths are determined by the pattern of the instantiation of
fundamental properties and natural relations. Since all objects involved in fundamental properties will be
concrete, A3 effectively confirms something like Chunky. Semantically speaking, A3 effectively

136

requires that the verification set V only includes occurent events involving fundamental properties and
natural relations.
T63: Hardcore Actualism: .( ee((ee ^ ee) ^ ( ee)))
If is true then its being true is grounded in some plurality of actually obtaining
fundamental events. This effectively means that the possible is reductively grounded in the actual, in
which is also grounded the dispositionl properties of objects. I call T78 hardcore actualism after the
similar doctrine advocated by Contessa (2010):
Hardcore actualists think that what makes modal propositions true are irreducibly modal features
of the actual world (such as laws of nature, dispositions, or essences) (Contessa 2010: 341-42).
I shall go into the consequences of hardcore actualism in the next chapter. First, however, I would like to
look at some additional consequences of A3.
T64: .x x
1. .( ee((ee ^ ee) ^ ( ee))) A3
2. ( ) ee((ee ^ ee) ^ (( ) ee))) 1. T [ ]/
3. PL
4. ee((ee ee) ^ (( ) ee))) 2, 3. PL
5. ee(ee ee) eee(e ee) ^ e)) Def. ee, CQT
6. ee(ee ee) 4. CQT, PL
7. e(e ee) ^ e) 5-6. PL
8. e e 7. CQT, PL
9. !e(e x(ex ^ x)) Def. x
10. ee x(ex ^ x) 9. CQT
11. x(ex ^ x) 8, 10. PL
12. .x x 11. CQT, PL, NEC
There necessarily exists some fundamental object. This does not, however, mean that there is anything
that is necessarily fundamental.
T65: Fundamental Base: .ee(ee)
By T79 and ee2.
Every possible world contains a fundamental base.
Do29: Metaphysical Nihilism: !xCx

137

Metaphysical nihilism is the doctrine that there is some possible world w, call it w, at which there are no
concrete objects. This would also, of course, mean that everything is such that it is possibly not concrete
!xCx.
T66: .x Cx
1. x x T79 T
2. x X(X ^ Xx) Def. x
3. X X Def. X
4. (X ^ Xx) (X ^ Xx) 3 PL
5. X(X ^ Xx) X(X ^ Xx) 4, CQT
6. X(X ^ Xx) Cx Def. C
7. x Cx 2, 5-6 PL
8. xx xCx 7, CCQT
9. .xCx 1, 8 PL, NEC
First Corollary: The Denial of Metaphysical Nihilism: !xCx
Second Corollary: x Cx
The fundamental level is not possibly empty. Since the class fundamental objects is a sub-class of the
class of concrete objects, it is also impossible that there should be no concrete objects at all. So we now
know that something or other is concrete! It is, however, to be admitted that A3 practically presupposes
that there are concrete objects.
4.2 On the Grounding of Abstracta
T67: Grounding Abstracta: !x(((x) ^ Ax) ee((ee ^ ee) ^ (((x) ^ Ax) ee)))
Any truth about an abstract object is either factual equivalent to or grounded in some fundamental facts.
Again, this is just an instance of the A3 schema. In the case of factual equivalence, we might have
something close enough to nominalism.
This theorem might strike the reader as an awfully strong demand to make of the concrete world.
This is because it is difficult to conceive what sort of states of affairs might be up to the task. One might
conclude that I should have followed Lewis in restricting A3 to contingent formulas:
( ^ .) ee((ee ee) ^ ( ee))
In that case, however, P would not permit the possible to be grounded in the actual, since by S5
. and so A3 would not require that modal truths should be grounded in non-modal truths.

138

Grounding Abstracta is perhaps not as strong as it seems. First, if there are no abstracta, then it is
perfectly vacuous. The present work, however, has already committed itself to abstract possible worlds
and the semantics seem to bear commitment to the existence of sets.
Nevertheless, I think that despite its strength, Grounding Abstracta should turn out to be true.
That is, it seems to me that the main problem nominalists have with abstracta is precisely the view that
abstracta cannot just exist in their own right; they should be dependent entities and dependent upon the
concrete. But given that abstract objects exist necessarily if at all (especially given NNE) their being
grounded in the concrete will place significant constraints on how the concrete could possibly be. This is
all basically true. I would like to suggest what options we may have here. I shall not endorse any one
particular strategy but only wish to show that there are some options on the table and that the choices are
not that awful. Moreover, I for one think that theorems that place unexpected restrictions on the possible
configurations of concrete reality are of particular interest. Such consequences are to be welcomed if we
take metaphysics seriously as a science and not merely a form of conceptual analysis.
First Strategy: Equating the Abstract with the Structure of the Concrete. Lewis (1991, 1993) is
exemplary of one sort of strategy that one might take, whereby mathematics is mostly reduced to
mereology. The reduction, however, requires that the cardinality of concrete mereological atoms should
be inaccessible, and necessarily so. The concern here is that we do not know if there really are
inaccessibly many concrete objects. This megethology is the result of supplementing classical
extensional mereology with plural predication.
We might call this the structural approach. Such an approach will not typically posit abstract
objects as such but will seek to equate truths about so-called abstract objects upon certain structures
exhibited by the concrete universe. In the case of megethology, the only structure required is cardinality.
Moreover, for the modal realists, propositions can be understood as sets of possible worlds. Indeed, this
approach probably requires modal realism, since one might also help oneself to the whole of concrete
reality when seeking ones mereological atoms and, given Recombination, it is not implausible that there
might be so many atoms. I cannot help myself to this solution. Not only am I strongly inclined toward
actualism, but by NNE all of the same entities must exist in each and every possible world. Hence
megethology will be a fairly implausible solution given the metaphysical system under discussion.
Second Strategy: Equating the Abstract to the Thoughts of an Infinite Intellect. Historically, the
most common way to identify abstracta with some sort of concrete process was by way of what one might
call divine conceptualism or infinite intellect conceptualism. On this view, held by the likes of
Plotinus and St. Augustine, abstract objects are conceptual objects, but they are not grounded in the
cognitive capacities of finite creatures so much as on the cognitive capacity of some divine entity with

139

infinite cognitive capacities. The intellect (or intellects) in question must be infinite in that its cognitive
powers must be able to ground completed irrational numbers, infinite choice sets etc. Menzel (1987)
proposes a sketch of a formal theory roughly indicating how this might work. If such truths are to be
necessary truths, then it must necessarily be the case that such an intellect exists (although it need not be
the case that anything in particular must be such an intellect).
Third Strategy: Equating the Abstract with Possibile Cognition. I would like to suggest a third
approach whereby all that abstract objects need in order to exist is to be cognizable, but they need not be
cognized. On the face of it, this proposal is problematic, since something has to already exist in order to
be cognizable. But perhaps the basic idea here can be stated in a clear and non-paradoxical manner.
Evoking possible worlds, it might be said that if x is an abstract object then there is some possible world
w in which x is clearly and distinctly conceived by some y. But have I not already characterized possible
worlds as abstract object? Am I not thereby evoking the abstract to ground the abstract? Not quite. Given
Hardcore Actualism, modal truths are in no way grounded in or made true by possible worlds, rather they
are made true by facts about the actual world. Accordingly, all that is being said is that an abstract object
x must be such that there is some y that has some iterated disposition to conceive of x. This strategy seems
to rest on some presumed similarity between the intentional nature of dispositions and the intentional
nature of mental states, and hence differs from the strategy above by suggesting that the intentionality
of dispositions is good enough to give us the existence of abstract objects. The putative fact that the
dispositional properties point beyond themselves gives us abstract objects consisting of all the ways
things might have been, including all the ways in which things might have thought or conceived of things.
Extensionalists like Quine and Lewis will have already gotten off the train as few stops ago. What
about everyone else? The key difference between myself and my chief rivals, namely the Humean
tradition, is that I understand the actual world to possess a kind of depth while my rivals deny that this is
the case. My rivals defend an ontology of chaos (i.e., nothing is genuinely caused) while I defend an
ontology in which things point beyond themselves (i.e., an ontology of powerful particulars); neither is
altogether free from features that may give rise to some degree of embarrassment but I prefer the latter to
the former.
Fourth Strategy: Equating the Abstract with Possible Concrete Structures. One might opt for
something closer to Lewis proposal. An abstract object x is grounded by the fact that the world, or some
part thereof, might have possessed a certain structure, which structure might have been characteristic of x.
The abstract object is what you get when you abstract the structure from the concrete features of such
possible states of affairs. For instance, structured propositions such as Fab would be grounded by (1)
the existence of a and b, and (2) the fact that, for some x, y, Fxy is true. That is, even if nothing

140

instantiates the F relation, some pair of objects might have. Given NNE, however, this sort of
modalization, would fail to make Lewis megethology more flexible in the case of sets.
Of these proposed solutions my favored solution is the possible cognizer solution. If is some
truth involving an abstract object x, then there is some y such that might have been conceived by y.
Such a theory would be in need of fleshing out. In particular, it will be necessary to become clear about
just what it means for a mind to conceive of something in general. Must every detail of an abstracta
be individually specified in order for it to be conceived, or need one merely produce a rule by which it
might be specified in principle? We ourselves only do the latter when defining most infinite sets. But how
do we get choice sets that cannot be obtained in virtue of repeated applications of a rule? And, while we
can define classes of choice sets, inaccessible sets, nonconstructable sets, etc., it is rare that we are able to
characterize individual members of these classes (on the assumption that they are not empty). If there is
no solution to these troubles, it might turn out that this strategy will require that there is some y such that
y is possibly an infinite intellect. If so, the present strategy would collapse into the infinite intellect
strategy should it be impossible for something to contingently fail to be an infinite intellect.














Copyright Adam Labecki 201

141

CHAPTER EIGHT:
HARDCORE ACTUALISM AND BRANCHING WORLD-STAGES

The first section of this chapter defends HCA by arguing that the range of possibilities that can be
grounded in the powers of actual individuals is appropriate for metaphysical modality. The second section
shows how possible words can be obtained by HCA modal truthmakers by construing possible worlds as
possible histories of the actual world. The third section introduces two axioms governing the modal
features of individual substances and presents several resulting theorems.
1. General Defense of HCA
1.1 General Background
While a number of authors have endorsed something along the lines of HCA in recent years,
106

this approach to metaphysical modality has generally been shunned or ignored. Perhaps the two most
serious objections to HCA in the literature are raised by Pruss (2002) and Cameron (2008). First, it is
claimed that hardcore actualism, at least when it is based on dispositions or laws, can, at most, provide an
adequate account of nomological, or physical, possibility. Indeed, it is commonly supposed that
dispositional essentialism (which was introduced in Chapter Seven) entails nomological essentialism,
being the doctrine that the laws of nature are the same in all possible worlds. Since it is generally believed
that conceivability is a fallible, yet reliable, guide to metaphysical possibility, the possibilities afforded on
hardcore actualism will be a small subclass of the class of metaphysical possibilities
.
Second, it is
suggested that, while hardcore actualists may be able to account for local possibilities such as Humphrey
might have won, it will be difficult to get global possibilities, or possible worlds, from dispositions or
laws alone. Should this be so, hardcore actualists might not even be able to help themselves to ersatz
worlds. I shall respond to each of these criticisms in turn.
I respond to the first charge in two phases. First, I show that hardcore actualist modal truthmakers
are capable of grounding possibilities that extend well beyond what philosophers typically mean by
nomological or physical possibility. Accordingly, I make a distinction between narrow and wide
nomological possibility. If all modal truthmakers are physical, metaphysical possibility would be co-
extensive with wide nomological possibility. If, however, concrete reality includes something like the

106
Theorists who might be called dispositional hardcore actualists include Martin and Heil (1999), Ellis (2001), Bird
(2006), Borghini and Williams (2008), Jacobs (2010), and Mumford and Anjum (2011). Pruss (2002, 2011)
proceeds in a similar manner, but advocates the traditional view that God is the ultimate truth-maker of metaphysical
possibility. Fine (1994) and Correia (2006) would appear to be a hardcore actualist of the essence-based sort.

142

God of theism, then metaphysical possibility will be considerably greater. (Or concrete reality might
include more than what we would regard as physical but does not include anything like God).
Second, I argue that metaphysical possibility is not as permissive as it is generally believed to be,
at least if one believes that metaphysical modality cuts nature at its joints. More specifically, I present an
argument from analogy to the conclusion that metaphysical possibility should be relatively sparse (as
opposed to abundant) and only discernible a posteriori. Nevertheless, I offer a consolation prize for
conceivability, namely that it may be a fallible yet reliable guide to meta-metaphysical possibility. From
these first two phases, I conclude that the identification of metaphysical possibility with wide-
nomological possibility is both plausible and attractive.
I respond to the charge regarding possible worlds by showing how ersatz worlds can be grounded
in HCA truthmakers by modeling such worlds by means of an adaptation of probabilistic branching-time
logic. This formalization, however, will be wholly semantic and is not integrated into the object language
of the theory. Such integration awaits a future project.
In modal epistemology is it often presumed that conceivability is a reliable guide to
metaphysically possibility.
Do30: Conceivability-Possibility (CP) is the doctrine that, for any proposition p p is
conceivable is a reliable basis for the inference p is metaphysically possible.
CP goes at least as far back as Descartes who, in Meditation Six reasons that anything he can clearly and
distinctly understand is capable of being realized by Gods power. Indeed, historically speaking, highly
permissive accounts of metaphysical possibility probably originate on the basis of belief in an omnipotent
God. Such permissiveness, however, is retained by Hume:
Tis an established maxim in metaphysics, that whatever the mind clearly conceives includes the
idea of possible existence, or in other words, that nothing we image is absolutely impossible
(Hume, 1739/2000: 1.2.2.8).
Indeed, Hume is the primary source for the contemporary acceptance of CP among analytic philosophers.
It should be noted, however, that Humes permissiveness about possibility seems to be symptomatic of a
deflationary stance toward metaphysical modality in general. It is reasonable to be unrestrictive about
possibility if one believes that modality is effectively a feature of how we think about things.
107

CP is somewhat weaker than Humes principle. Again, it is generally believed that conceivability
is a reliable guide to possibility. Accordingly, if p is conceivable, the burden of proof falls on anyone who

107
Cf. Hume (1739/2000: 1.3.14.22)

143

would deny that p is possible. That is, possibility claims are presumed true unless shown to be false. CP is
typically presumed in most discussions on modal metaphysics.
But why should we believe CP? I cannot here go into every plausible reason for holding CP
108
but
I concur with Wilsons (forthcoming) conclusion that the main reason for adopting CP is due to the
apparent lack of alternatives. Whatever the range of metaphysically possibility may be, it should be more
restrictive than mere logical possibility, which requires nothing more than logical consistency, but less
restrictive than nomological possibility, which requires that we keep the effective laws of nature fixed. It
is difficult, however, to find a principled way to discern just which possibilities are genuinely
metaphysical. I believe that CP is commonly held because it is tacitly believed that we cannot do better
and hence that its denial will lead to modal skepticism.
109
If, however, there is a principled way to discern
a modality whose extension lies between the nomological and the merely logical, then denying CP will
not lead to modal skepticism. I shall now proceed to show that there is such a way.
1.2 Wide-Nomological Possibility
There are at least two respects in which the range of possibilities grounded by physical actualist
truth-makers exceeds those which are typically regarded as nomological. The first is due to the fact that
the actual laws of low-energy physics are the result of contingent historical processes. The second is due
to the fact that quantum mechanics (QM) assigns minute, but non-zero, probabilities to some pretty
miraculous behavior (although, in truth, the former is probably reducible to the latter).
The laws which govern ordinary low-energy physics are physically contingent inasmuch as
cosmological theories have it that there are times in the distant past, when the universe was very small
and hot, at which the fundamental forces known to us were not yet separated from a prior unified force.
For instance, the symmetry between electromagnetic force and the weak force is believed to break when
the temperature of the universe has reduced enough for the W

and Z
0
bosons to acquire mass through
interaction with the Higgs field. Prior to this symmetry breaking W

and Z
0
bosons are indistinguishable
from photons (i.e., electromagnetic interactions were indistinguishable from weak force interactions).
It is believed that the symmetry-breaking process by which the four fundamental forces came to
be separated is highly indeterminate, meaning that at least the nature of those forces, as well as the 3+1
dimensionality of the universe, are the result of contingent historical processes. If the number of

108
Yablo (1993) provides a good overview of cases for and against CP.
109
In some ways this is misleading, if only because it might be said that CP presupposes modal skepticism. It is just
because we cannot perceive necessary connections between distinct objects that we abstain from positing any such
connections.

144

dimensions is contingent then so is the fact the world include physical aggregates, since physical
aggregation requires 3+1 dimensionality. Tegmark (2003) summarizes as follows:
The prevailing view is that the physics we observe today is merely a low-energy limit of a much
more symmetric theory that manifests itself at extremely high temperatures. This underlying
fundamental theory may be 11-dimensional, supersymmetric, and involve a grand unification of
the four fundamental force of nature. A common feature of such theories is that the potential
energy of the field(s) driving inflation has several different minima (sometime called vacuum
states), corresponding to different ways of breaking this symmetry and, as a result, to different
low-energy physics (Tegmark 2003: 265).
110

It might well be the case that the laws of high-energy unified physics are the same in every possible
world (although this is epistemically contingent upon future developments in physical cosmology). For
instance, presuming that the underlying nature of physical energy is the same in all possible worlds, the
conservation of energy would likely hold in every possible world.
111
Most laws of actual low energy
physics (i.e., most laws that describe the observable universe), however, are physically contingent.
In particular it should be noted that physical laws can effectively be reduced to the minima or
vacuum states mentioned by Tegmark above. If the cosmological constant has a positive value,
whereby it may explain the accelerating expansion of the universe, then it might be identified as this
vacuum state. If the cosmological constant is not positive valued, then it might be identified with or
quintessence. In either case it might be identified with dark energy.
112
If the minima were lower than its
actual state (i.e., if vacuum state was colder than it actually is) then we might have witnessed additional
symmetry breaking from the four existing forces. The main point, however, is that the properties of
individual particles (provided that particles are individuals) and the value of the global minima conjointly
determine the (probabilistic) laws of the actual world.
Despite this well-known result of the intersection of cosmology and particle physics
metaphysicians continue to write about the laws governing the Standard Model as if they characterized
the entire lifespan of the universe. Y. Balashov (2002), however, has taken note of this feature of
cosmology and has used it to defend dispositional essentialism by denying that it entails nomological

110
Lind (1985) was among the first to write on this matter: Without the scalar field there is no difference between,
say, the weak nuclear force and electromagnetism. But when an appropriate scalar field fills the Universe, it has the
effect of breaking the symmetry between these two forces, making them go their separate ways (Linde 1985: 15).
111
This would seem to entail that the universe can never naturally come to an absolute end (at most it might assume
a vacuum state in which no energy is in the form of a material particle).
112
A fundamental problem in QFT is that it results in a vacuum state that is 40 orders magnitudes greater than is
warranted by observation (which is a little less than 3 Kelvins) and which would only be expected to hold for the
very early universe (just after the Plank Era). In general, the observed value of the cosmological constant cannot be
adduced by particle physics. This, again, is one reason why physicists are seeking after a more fundamental and
unified theory of high-energy physics.

145

essentialism. We can know that the effective laws of the universe are contingent precisely because they
have actually been different in the past, if only during the initial moments of the universe.
For any dispositional ascription D
(cm)
xt, the value of the universal vacuum energy should be
included in c. Ordinarily, one fails to mention such ubiquitous features of the world, but they causally
contribute to the occurrence of m and so they must be included. Hence, if a dispositional property X
bestows D
(cm)
xt upon its bearers in the actual world, it does so in every possible world, but it just so
happens that in many worlds some objects are X and yet c never obtains. In such worlds we would have
alternative low-energy natural laws while the overarching laws of high-energy physics are likely to be
held fixed.
Accordingly, one way to reach relatively remote possible worlds is to consider ways in which the
world might have evolved from its very early stages, or, should there be no beginning of time, how the
world might have evolved from one of its hot and compressed phases.
A second source of remote possibility is due to the fact that, on most interpretations of QM,
physical systems often exhibit very small objective probabilities for highly uncharacteristic behavior. For
instance, while quantum tunneling through barriers is not terribly exotic for low mass particles, it has
probably never happened that a bullet simply passed through a mammal without causing any harm.
Nevertheless, there is always a negligible but non-zero probability for a bullet to tunnel through a
mammal rather than damage it (as well as itself). In all likely-hood such an event will never occur in this
world, but it could.
In the case of the Everett interpretation of QM, in which objective probability is replaced with
literal world-branching,
113
quantum probabilities of this sort seems to lead to something like quantum
immortality. For any given organism x, any (or nearly any) time t at which there is some probability of x
dying at t+1 is also a time at which there is some probability, however small it may be, of x living at
t+1.
114
On an ersatz interpretation of world-branching, this just means that for any organism x, there are
some ersatz worlds in which x has an astonishing long lifetime. Immortality is simply a salient instance of
a much larger class of possibilities introduced by quantum probabilities. There is, for instance, also a
minuscule chance that the particles that presently compose my bed will buzz about the North-West corner

113
More precisely, on the Everett interpretation objective probability is replaced with subjective probability in that
one does not know to which world one belongs.
114
See Tegmark (1998) for an early exposition of this result, which was apparently noticed by Everett himself.
Lewis (2004) has indicated that quantum immortality would be more like quantum torment, since most cases of
survival over non-survival will involve significant and lasting injury. For instance, while it could happen that
the whole bullet tunnels through ones body, it would be less improbable for most of its atoms to tunnel while some
fail to tunnel. In such cases said bullet would cause limited damage to a body. That is, most cases of improbable
survival would involve some degree of harm.

146

of my room next time I leave for the grocery store, only to be in a perfectly ordinary configuration upon
my return.
If the dispositions or laws of the actual world result in some objective probability, however small,
that a state of affairs S should obtain, then surely the hardcore actualist may claim that S is metaphysically
possible and said dispositions or laws make it the case that S is possible. Moreover, our justification for
believing that S is possible arises from natural science. In many cases, the present state of science will
permit us to discern whether S is possible, but the key is that, regardless of this, there will be a fact of the
matter as to whether the world might have evolved in such a way that S was the case.
Let us call those quantum events which are so improbable that they are unlikely ever to be
observed by humans, quantum flukes. Some worlds will be richer in quantum flukes than others;
indeed, in some worlds there will be individual objects involved in a whole series of flukes, such as
worlds in which Napoleon lives to be 400 years old. In other worlds there will be relatively small
spacetime regions in which a large number of quantum flukes occur. Worlds that contain such fluky
patterns might be called fantastic worlds. Fantastic worlds would provide the hardcore actualist with a
second class of remote worlds. If, however, the actual already world happens to be fantastic, then worlds
which are fantastic in ways in which the actual world is not fantastic will count as remote.
The hardcore actualist, then, can accept both worlds with different effective low-energy laws as
well as worlds in which the actual low-energy laws generate fantastic patterns of highly uncharacteristic
behavior. Indeed, a third class of worlds may be obtained by combining the first two: worlds with
different laws which are also fantastic relative to said laws. All of these possibilities are, strictly speaking,
physical, since all of them are grounded in (actual) physical reality. In the very least, however, one will
want to distinguish between narrow and wide nomological possibility. Narrow nomological possibility
restricts itself to the contingent laws of low-energy physics and excludes fantastic possibilities. When
philosophers speak of nomological possibility they typically have the narrow sort in mind. Wide
nomological possibility includes every set of possible low-level-energy laws and every pattern of
quantum flukes of non-zero probability. As such, the range of wide nomological possibilities is
considerably greater than the range of narrow nomological possibilities.
When restricting itself to naturalistic modal truthmakers, hardcore actualism would identify
metaphysical possibility with wide nomological possibility. My claim is that, if the only modal
truthmakers are of a broadly physical nature, wide nomological possibility would be wide enough for
metaphysical possibility. Given my agnosticism regarding physicalism I am also agnostic as to whether or
not metaphysical possibility is identical with wide nomological possibility but I take it that, given HCA,
the range of metaphysical possibility will be at least as wide as the range of wide nomological possibility

147

and that the latter is wide enough. Should there be additional modal truthmakers, in the form of gods,
entelechies, etc., then so much the better for HCA.
1.3 Against Abundant Metaphysical Possibility
While wide nomological possibility may include some worlds that are very remote from our own, we
can surely conceive of situations which are just not physically possible, even in the widest sense of the
term physical. Hence the range of possibilities grounded by HCA will fall short of the range of
possibilities obtained on the basis of CP. For instance, worlds in which there are populations of floating
human heads, but in which there have never been any human bodies, probably do not fall under the scope
of wide-nomological possibility. Likewise, it does not appear that there would be zombie worlds: worlds
in which there are beings with human physiology lacking consciousness. The laws cannot just vary in any
way one fancies. Contingent laws alter macro-conditions (in fact they are macro-conditions), they are not
fine turned to alter some species of phenomena while leaving others completely unaffected. Physics
teaches us that reality is not as loose and disconnected as Humeans seem to think. The parameters
permitting carbon-based life are very narrow. Indeed, the universe must be rather fine-tuned in order to
permit physical aggregation at all. While the laws might have differed in many ways there are very few
nomological-alternatives for familiar macro objects. Quantum flukes provide some non-standard
behaviors, but not just any non-standard behavior that one might imagine.
115

I have already argued that adherence to CP is optional provided that one can provide an
alternative. Nevertheless, I believe that many philosophers will continue to insist upon CP, if only
because the principle has been so deeply ingrained in the culture of analytic philosophy. I shall address
this by producing an argument aimed directly against CP and by indicating just where I believe CP is both
appropriate and important for the discipline of metaphysics.
As a result of the widespread acceptance of CP, it is generally thought that metaphysical
possibility is abundant in that it should satisfy some sort of plentitude or comprehension principle.
Abundance here is to be contracted with sparseness, as we saw in the distinction between sparse and
abundant properties. Metaphysicians like Lewis and Armstrong are primarily concerned with sparse, or
natural, properties because those are the ones that are supposed to cut nature at the joints. Abundant

115
Some theorists respond to arguments from conceivability by placing certain restrictions on what it means for
something to be conceivable. Such theorists tend to come to similar conclusion as myself. I do not go this route
for two reasons. First, it is difficult to precisely define conceivable in a plausibly restrictive way and doing so
involves subtle considerations from the philosophy of language and the philosophy of mind. Second, conceivability
does not appear to be a relevant to the consideration of metaphysical possibility.

148

properties are discerned in a more or less a priori basis (there is a property for each subset of the set of all
objects) while sparse properties are discerned a posteriori in light of natural science.
One can also make a distinction between abundant and sparse composite objects. On an abundant
conception of composite objects, there is a composite object for any fusion of any number of parts of
objects.
116
On a sparse conception, one wants to include only those composite objects that might rightly
be regarded as an integrated substance or, at most, objects which possess some principle of unity. While
there is a mereological fusion of the Moon, the top half of the Eiffel Tower, and Lewis left ear, such
objects appear to be as unnatural as the property grue, while some composites, in particular living
organisms, at least seem to cut nature at the joints. Again, abundant objects are discerned in a more or less
a priori basis (effectively in the same manner as properties) while sparse objects are discerned a
posteriori in light of natural science.
By analogy, then, when doing metaphysics we should not be interested in an abundant conception
of possible worlds so much as those (relatively sparse) worlds which cut at the modal joints of nature and
which are discerned on an a posteriori basis. I say relatively sparse, because there will still be far more
possible worlds than there are natural properties or instances of natural kinds, but they will be sparse
relative to the plurality of worlds one needs in order to verify CP. In particular, we should not regard any
general plentitude principle as a reliable guide to possibility, since any such principle will deliver
abundant worlds rather than sparse ones.
This little argument from analogy is by no means a refutation of CP as applied to metaphysics. Its
purpose is merely to indicate that there are positive reasons to suppose that CP is an inappropriate
principle to follow when doing metaphysics.
1.4 Conceivability and Meta-metaphysical Possibility
As a consolation prize, and something of an olive branch, to the advocates of CP, I suggest that
what has frequently passed for metaphysical possibility does answer to a certain kind of possibility and,
moreover, a kind of possibility that is relevant to metaphysics as a discipline. One might desire an
abundant conception metaphysical possibility because one believes that such is required in order to
compare and evaluate competing metaphysical theories. I think that the proper modality for such a task is
not metaphysical but meta-metaphysical, since it is precisely the sort of modality one employs when
doing meta-metaphysics.

116
Where, for any object x, there is generally thought to be an isomorphism between parts of the space occupied by
x and the mereological parts of x. This view often leads to an identification of objects with mereological sums of
spacetime points, where there is a mereological sum for any collection of spacetime points whatsoever.

149

Lewis (1986a) sometimes seems to have this meta-theoretical role in mind when elaborating his
possibilist theory of modality but I believe that it was folly to suppose that such a wide theory of
possibility requires substantive truthmakers. That is, in the case of meta-metaphysical possibility, facts
about our cognitive capabilities seem to be just the modal truth-makers we need. The need for meta-
metaphysical possibility arises simply from fallibilism regarding our knowledge of the nature of the actual
world. Relative to the most stringent epistemic standards none of us really know which theory of the
world is right (indeed, we do not know if metaphysics is even humanly possible), and so none of us know
which possibilities are grounded in the actual world. Hence we need only add an epistemic dimension to
our analysis.
I hereby suggest that CP is a fallible but reliable guide to what one might call meta-metaphysical
possibility. The distinction I draw between meta-metaphysical and metaphysical possibility is similar to
the distinction Chalmers (1996, 2010) makes between primary and secondary possibility. It makes use of
two-dimensional analysis
117
whereby propositions are evaluated relative to pairs of possible worlds,
where the first world is taken to be epistemically possible while the second is metaphysically possible
relative to the first. A world is epistemically possible when it is possible for all we really know. In
this context we are relatively skeptical regarding which theory of the world is true. That is, we suppose
that any theory that is worthy of consideration could be true for all we know.
The idea is that the range of metaphysical possibility is epistemically contingent upon the way the
world actually is. On a two-dimensionalist analysis one might make claims of the following form: if the
actual world is a p-world then every world is a p world or if the actual world is a p-world then there is
some q-world. Now, it is also the case that which theories about the world are true depends upon what
the world is actually like. To each metaphysical theory or doctrine one can assign a subset of
epistemically possible worlds, W. Each w e W is such that, if w is the actual world, then can be
regarded as an adequate theory of the world.
118
I would like, then, to propose the following definition of
meta-metaphysical possibility:
Df82: Meta-metaphysical Possibility: A proposition p is meta-metaphysically possible just in
case there is some reasonable and informed theory , verified by some epistemically possible
world w, such that, should w = w
@
, then there is some metaphysically possible world w such that
p is true at w.

117
Two-dimensionalism has been famously employed by Chalmers (1996) and Jackson (1998). My own use of two-
dimensionalism, however, does not place especial significance to what certain terms would designate between
distinct epistemic worlds.
118
Nearly any epistemically possible world w will verify more than one theory , in which case theories must be
evaluated in terms of their theoretical virtues, such as parsimony.

150

When we are relatively liberal regarding which theories are both reasonable and informed (and I am
supposing that we should be), nearly all propositions that are conceivable will turn out to be possible and
hence conceivability will, indeed, turn out to be a reliable guide to meta-metaphysical possibility.
That metaphysical possibility should often be identified with meta-metaphysical possibility is not
surprising, since metaphysicians spill much of their ink doing meta-theory on their own subject matter
and it is sometimes difficult to draw a clear line between metaphysics and meta-metaphysics.
Accordingly, the sort of possibility that has been taken to denote metaphysical possibility may still be
employed provided that it is used to make claims about metaphysical theories or doctrines rather than
claims about the nature of reality as such. I suspect that cases in which arguments from metaphysical
modality are liable to raise the most suspension are precisely those which attempt to derive substantial
claims about reality on the basis of meta-metaphysical possibilities. One may discern this to be the case
when Chalmers (2010) attempts to refute mind-body physicalism on the basis of the conceivability and
hence metaphysical possibility of zombies. When doing meta-metaphysics, however, one may argue that,
should physical monism be true, zombies are metaphysically impossible. Likewise, any theory according
to which zombies are possible is a theory according to which physicalism is false. What we should not
conclude is that physicalism is false just because zombies are conceivable.
119

Moreover, even for the hardcore actualist, for all we know, it could be the case that CP holds for
metaphysical possibility. It may, for instance, be satisfied if theism is true.
120
One will be justified,
however, in holding CP only inasmuch as one is justified in believing that the actual world is sufficiently
robust in its modal truthmakers to render metaphysical possibility nearly as extensive as meta-
metaphysical possibility.
There might well be a general correlation between ones interest in meta-metaphysics and ones
views on metaphysical possibility. That is, one might find that those who are strongly attached to CP as a
guide to metaphysical modality are already using it for meta-metaphysical purposes most of the time. For
instance, Sider (2001) suggests that the best way to defend a metaphysical theory is show that it is better
than all of the alternatives on the table the last theory standing wins. This methodological approach
encourages a style of metaphysics that is more focused on meta-theory than object-theory. Likewise,
those who believe CP is too permissive to be a reliable guide to metaphysical modality may tend to be
those, like myself, who are primarily interested in theory building, where theory comparison is a
secondary concern.

119
Nor should we conclude that dualism is false just because it is conceivable that the mind might be identical with
the body.
120
Pruss (2002) advocates a theistic form of hardcore actualism in order to retain CP; or perhaps he trusts in CP
because he is already a theist.

151

2. Hardcore Actualism and Branching Worlds
2.1 Motivation
HCA is largely motivated by the judgment that metaphysical possibility should be limited to what Kant
called real possibility. For a possibility to be a real possibility it is not enough that it should merely be
free from contradictions, it must be positively grounded in something actual. That is to say, expresses is
a real possibility if and only if there exists something in the actual world with the power to make it that
case that .
Metaphysicians have typically supposed that possibilities do not need to be grounded on anything
besides mere consistency. But this is only justified inasmuch as one is speaking of properly abundant
modalities, such as logical or conceptual possibility. Metaphysical modality should have something
substantive to say about the nature of reality in itself, not merely the nature of logic or about human
cognition.
Consider the conceptual possibility of a world w in which there were no concrete existents and
hence nothing with fundamental properties. According to HCA, it would follow that w cannot see any
worlds other than itself. That is to say, from the point of view of w nothing concrete is metaphysically
possible! This is because nothing would possess any dispositional properties and all mere possibilities are
to be grounded in the fundamental dispositional properties of things. Possibilities themselves must be
grounded and if there were nothing to ground them then nothing alone would be possible.
Given HCA one will look to features of this world when considering how the world might have
been. The HCA attitude toward modality does not concern how some world or other might have been
but rather how the actual world might have been. What does it mean, however, to say that the actual
world might have been in such and such a way? Could the actual world have consisted of nothing but two
iron spheres? I suggest that, in asking how this world might have been, we should consider how it might
have evolved from some past time.
2.2: Branching-Stage Logic
One way to model the ways in which the actual world might have been involves making use of
nonlinear temporal logic.
121
Traditionally, branching-time logic understands possibility relative to some
privileged present time t
@
, where is possible at t
@
just in case there is some t lying in some future of t
at which is true. H. Deutsch (1990) denotes such possibilities as real possibilities, since they are

121
Thomason (1984) has become the standard introduction to branching-time logic.

152

possibilities that remain open relative to the present. Real here is only relative to the present. When we
take up the timeless point of view appropriate to metaphysics, we should forego this restriction to the
present. When doing so, it turns out that is possible just in case there is some time t, perhaps a very
early one, such that is true at some t, where t lies within some possible future of t. As Mller (2012)
has independently observed, the resulting modality is broad enough to count as a general modality for
possibilities which are real in that they are grounded in the concrete universe. Modeling possibility in
this way will yield precisely the ways in which this world might have been. Barbarossa might have been
a successful campaign certainly seems to be true. Current events, however, cannot ground this truth,
since there is no possible future, relative to the present, in which Operation Barbarossa is successful. We
need to go back at least to sometime in the early 1940s if we are to come to a time for which the a
successful Barbarossa lies in a possible future. Likewise, the truthmakers required for the (low energy)
Laws of Nature might have been different may lie in the very earliest moment of the Big Bang.
In order to explain metaphysical possibility in this way it is required that we either accept
objective probabilities, which are correlated with the indeterminism exhibited in the behavior of the
world, or in a many-world interpretation of QM. I shall generally cater the exposition to the former
possibility.
122

The actual logic to be employed, however, will not technically involve branching times. This is
because every time necessarily exists in that T is no less constant between possible worlds than D. This
difficulty can be remedied easily enough be replacing branching times with branching world-stages.
Definitions 85-87 introduce the technical distinctions required to accomplish this replacement. Aside
from these differences the resulting branching stage logic is structurally identical with standard branching
time logic. Again, the presentation of branching state logic (BSL) will be wholly semantic, its full
integration with the semantics and object language of the theory being a project for the future. It will,
however, be shown that the resulting modality verifies the axioms of S5.
Df83: w-State Description: For each world w, let
w
be a subset of such that, for each e, [e]
, [e]
w
if and only if [e] = r
n
, [
1
]
M
, [
n
]
M
,

[t]
M and [
1
]
M
, [
n
]
M
, [t]
M
e
I (r
n
)
(w).

w
includes just those simple events which occur in w.
Df84: For every time t. let
t
be a subset of such that, for each e, [e]
M
, [e]
M

t
if and
only if t [e]
M
.
t
is the set of events that occur at t, at some world or other.

122
Something like an ersatz view of the Everett interpretation of quantum mechanics is widely held by physicists as
an obvious truth. Whether or not Everett himself was a realist about branching worlds is debatable, although some of
his followers, such as D. Deutsche (1985), maintain a realist stance on branching worlds.

153

Df85: Let
TW
be a set of subsets of such that, d
TW
if and only if there is some w W
and some t such that [t]
M
e T such that, for any e, [e] d if and only if e
w
and e
t
.
Each member of
TW
is effectively a tw-state description, for some t, w. That is, each member of

TW
corresponds to a world-stage: a complete description of how some world is at some time. Such tw
state descriptions effectively take the place of the times of branching time logic. I shall use d, with or
without subscripts, as a variable for members of
TW.

Just as events inherit their temporal ordering from the times they involve, world-stages inherit a
temporal ordering from the simple events they include (each such event involving one and the same time).
Df86: Tree Frame: A tree frame is a pair
TW
, < where
TW
is the set of all possible world-
stages and < is a relation over members of
TW
subject to the following conditions:
(1) < is an irreflexive, transitive, and backward-linear ordering (i.e., for all d
1
, d
2,
d
3
e
TW
, if d
2
< d
1
and d
3
< d
1
then d
2
= d
3
or d
2
< d
3
or d
3
< d
2
).
(2) For any d
1,
d
2
e
TW
, if d
1
< d
2
, then there exists some d
3
e
TW
such that d
1
< d
3
and there is
no d
4
e
TW
such that d
1
< d
4
< d
3
.
(3) For any d
1,
d
2
e
TW
, if d
1
< d
2
, then there exists some d
3
e
TW
such that d
3
< d
2
and there is
no d
4
e
TW
such that d
3
< d
4
< d
2
.
(4) For any d
1
, d
2
e
TW
, there exists some d
3
e
TW
such that d
3
d
1
and d
3
d
2
.
(Where is just like < except it is reflexive)
Condition (1) permits < to simulate the branching structure characteristic of branching-time logic only the
branching here is in terms of world-stages rather than times. Accordingly, it will not be the case that every
pair of stages will be related by <; in particular, when there is no world w such that d
1
and d
2
are both
stages of w, d
1
and d
2
will not be related by <. Condition (2) and (3) stipulate that time is to be regarded as
discrete. Condition (4) guarantees that every pair in
TW
is rooted. The condition is intuitive given
hardcore actualist assumptions, in that it permits us to model precisely how this world might have turned
out. As Mller (2012) notes, its presence will be necessary for defining general, and hence
metaphysical, modal operators within a branching-time structure. The branching spacetime counterpart of
(4) is sometimes called the prior choice principle (Belnap 1992); I shall refer to it as PCP.
Df87: For any d, d e
TW
, if d < d, then d is an ancestor of d and d is a descendent of d.
Df88: For any tree-frame (
TW
, <) and d e
TW
, a sub-tree
TWd is a subset of
TW
ordered
by < and having d as its minimal member under <.

154

By PCP, for any d
1
, d
2
e
TW
, there exists some d
3
such that d
1
, d
2
e
TWd3
. Also, we may say that for
any d e
TW
,
TWd minus d is the set of all possible futures (or descendants) of d.
Df89: Every linearly ordered subset of
TW
is a chain (or path) on
TW
.
Df90-: Let h be a function such that, for each w W, h(w) is a non-empty maximal linearly
ordered subset of
TW
. For each w, h(w) is the history of w.
123

In other words, for each w, h(w) is a partition of
w
relative to times. Hence, each world w will not only be
associated with a subset of but with a maximal linearly ordered subset of
TW
, although the former is
identical with the union of the latter. In terms of tree-structure, world-histories are like branches and are
often named as such. I shall let H denote the set of all world-histories. The non-emptiness requirement is,
of course, justified by NE, which also permitted us to deduce that non-vacuity of times.
T68: H/< Downward Closure: For any d, d e
TW
and h(w) e H, if d e h(w) and d d, then
d e h(w).
This is an immediate consequence of backward linearity and Def. 92.
Df91: For any d e
TW
, d is a root in
TW
, written d, just in case there is no d e
TW
such
that d < d.
A root would be a minimal element of
TW
and would effectively require that there be a minimal
member of T.
T69: There can be no more than one root in
TW
.
Suppose that d and d are distinct roots in
TW
but PCP is nevertheless true, that is, for any d
1
, d
2
e

TW
, there exists some d
3
e
TW
such that d
3
d
1
and d
3
d
2
. Hence, there must be some d such that d
d and d d. Since both d and d are minimal, d d and d d can only be true if d = d and
d = d is true. But, by transitivity of identity, it follows that d = d, which contradicts our initial
supposition. By reductio, the supposition must be rejected, and hence there cannot be more than one root
in
TW
.
In the absence of PCP
TW
, < could support multiple roots and hence forests, where forest
indicates that there are effectively multiple trees with null intersections. Forests would include worlds
which are in no way related to our own, and hence worlds that lack substantial actualist truthmakers (they
would be at most conceptually or epistemically possible).

123
D maximally linearly ordered subset of
TW
if and only if D is linearly ordered subset of
TW
such that if d e

TW
and d D, the union of D and {d} is not linearly ordered.

155

T70: If there is a root, d, in

TW
then, for each w e W, d e h(w).
Suppose that there is a d, which we know to be unique by T84. By T83, whenever a stage d belong to a
history h(w) so does every earlier stage. Hence, for any d, d e
TW
, and any w, w e W, it follows that d
e h(w).
Whether or not there is a beginning of time turns out to be a necessary matter; if the world has a
beginning then every world begins in the same way. Hence the necessity of origins is conditionally true
for both the world and all objects, since everything would exist at d and possess it necessary origin
therein.
If there is no minimal element in
TW
then a stage d could only belong to every possible world if
there were an infinite number of stages prior to d in which no branching occurred. If, however, we believe
in objective probability roughly in line with QM, every stage d would have more than one possible
successor and hence, should there be no root on
TW
, there would be no single stage common to every
possible world. Worlds, however, remain connected in that, for any two worlds there is some initial
segment that they hold in common. When talking about the class of all world histories, however, it must
be that case that, for any time t, there will always be some t, t < t, such that some worlds diverge at t.
Df92: For every d e
TW
, H
d
is the set of world-histories such that d h(w).
For any d e
TW
, H
d
will contain a single linear past prior to d as well as all of the sub-branches that
find their origin in d. Hence H
d
is effectively
TWd
augmented by the set of the ancestors of d, the result
being the set of all worlds of which d is a member.
Df93: Where h(w) and h(w) are elements of H and d e
TW
, h(w)
d
h(w) is true if and
only if for every d e
TW
, if d d, then d e h(w) h(w).
That is, h(w)
d
h(w) tells us that the histories of w and w are identical at least up to (and including) d.
T71: For every h(w) and h(w) H, there exists some d e
TW
such that h(w)
d
h(w).
For reductio, suppose there exists some h(w) and h(w) H such that there is no d e
TW
such that h(w)

d
h(w), and so h(w) h(w) is empty. Hence for any d, d e
TW
such that d e h(w) and d e h(w),
there is no d such that d d and d d, since, by downward closure, any such d would belong to the
intersection of h(w) and h(w) which, by supposition is empty. But this contradicts PCP, and so the initial
supposition must be denied.

156

The following set of definitions and resulting probabilistic branching-stage model is largely an
application of Mller (2005, 2010) to the present framework, which in turn, is a straightforward
application of basic probabilistic logic to branching logic.
Df94: For every h(w) and h(w) H, and d e
TW
, d is the point of divergence for h(w) and
h(w), written h(w)
d
h(w), just in case d is maximal in h(w) h(w) under <.
When d is the point of divergence for h(w) and h(w), we say that h(w) and h(w), branch (or diverge)
at d. One way to compare world-histories for similarity would be in terms of their point of divergence.
That is, if h(w)
d
h(w) and h(w)
d h(w), then h(w) is more similar to h(w) than h(w) just in case d
< d. While such a method of comparison is extensionally straightforward there are no doubt potential
counterexamples, i.e., worlds that branch late but dramatically versus world that branch a littler earlier but
in subtle ways.
Df95: For any d
1,
d
2
e
TW
, and h(w) H, d
2
is the successor of d
1
in h(w), written d
2
= s(d
1
)
w
,
just in case d
1,
d
2
e h(w), d
1
< d
2
and there is no d
3
e
TW
such that d
1
< d
3
and d
3
< d
2
.
A successor is an immediate descendent and Def. 76, clause (2), guarantees that any non-maximal
moment d has some successor. Any d e
TW
can have at most one successor in any given world w, but,
provided that d is a point of divergence for distinct worlds, it will have several distinct successors in H
d
.
Df96: For any d e
TW
, S
d
= {d| d = s(d)
w
, for each h(w) H
d
}.
For any stage d, S
d
is the set of possible successors of d. The members of S
d
may be called the children
of d and d the (single) parent of each member of S
d
. Likewise, pairs of members of S
d
can be called
siblings. In terms of probability theory, S
d
is a sample space. It will be assumed that S
d
is finite, since
probably theory involving infinite probability spaces is fairly sophisticated
124
.
Df97: For any d e
TW
, P is a function assigning each member of S
d
to a member of .
Df98: TW Probability Model = W, D,
TW
, H, h(w
@
), <, , S
d
, I , J

TW
, H, <, , and S
d
are as defined above while W and D are as defined in Chapter Two.
h(w
@
) is distinguished member of H, denoting the history of the actual world.
I whose application to each n-place predicates r effectively determines which events occur at
each member of
TW
. That is, in determining the intension of each predicate extension, I
determines which events are members of
w
for each w, e W (by Df. 85).
w
, in turn, determines
the content of h(w) of each member of H (by definitions 86-87 and 92).

124
See Williamson (2007) for some troubles involving infinite sample sets, along with some possible solutions.

157

J is a probability function such that, for each d e
TW
, J (d) is a function P from members of S
d

to members of subject to the following constraints:
(1) For all d, if d e S
d
then P(d) > 0 (Positivity)
(2) For all d, P(S
d
) = 1 (Normalization)
125

Since, by A2, we are assuming that the total categorical state of a world-stage is equivalent with its
dispositional state, it is assumed that the interpretation assignment I effectively determines the J function.
That is, for any I valuation there is a unique J valuation.
126

Finally, the variable assignment will be extended to assign each stage variable d to a member of

TW
and each r variable to a member of .


The truth valuation for formulas regarding the probability of one world-stage succeeding another
is as follows:
[VP(d)]: V
M
(P(d) = r, d, w) = 1 iff d e S
t
and P: d r e J (d)
Unlike the propensities had by many, if not all, individual objects, the propensity of a stage d will not be
conditional upon some external condition, since we are talking about the entire world at a given time.
Indeed, this permits a direct characterization of the total propensity (or dispositional state) of the world at
a given stage d in terms of standard, non-conditional, probability spaces.
127

Having sketched a probabilistic branching-time logic, I now introduce an extra-logical postulate
which is intended to make the connection between actualist truth-makers and possibile worlds more
intuitive and precise.
PC-Postulate: For any d
1
, d
n
, where d
1
e h(w
@
) and d
n
e
TW
, d
1
makes d
n
possible if and only
if there exists some chain d
2
d
n-1
such that d
2
e S
d1
, d
3
e S
d2
, d
n
e S
dn-1.
In other words, d
1
d
n
if and only there is a non-zero probability of d
n
eventually occurring in virtue of
the actual occurrence of d
1
. When this is the case, d
1
possesses the non-conditional iterated disposition to
manifest d
n
. This is semantically expressed by the existence of a chain such as d
2
e S
d1
, d
3
e S
d2
, d
n
e
S
dn-1
, which

might be called a possible-cause chain, or a PC-chain. The PC-Postulate is effectively a
metaphysical cum causal interpretation of probabilistic branching stage logic. More precisely, the PC-
Postulate is introduced so as to guarantee that the probabilistic branching-stage logic will verify HCA,
which is required for the verification of A3.

125
In light of probability logic, the analysis of dispositional ascriptions might also be modified so as to include
probabilities. This would involve assigning a member of to m.
126
In principle, however, it would have been preferable to make the relation between the I and J functions explicit.
127
Surez (2010) has shown that absolute, or non-conditional, propensities bypass the primary obstacles that arise
when trying to correlate probabilities with propensities.

158

T72: For any d
1
, d
2
, d
n
e
TW
, d
2
e S
d1
, d
3
e S
d2
, d
n
e S
dn-1
if and only if d
1
d
n
is a
chain.
Left to right: Suppose d
2
e S
d1
, d
3
e S
d2
, d
n
e S
dn-1.
By Def. 87, for any d, d e
TW
, distinct form d, if
d e S
d
, then d < d. Hence, from d
2
e S
d1
, d
3
e S
d2
, d
j
e S
dn-1
we may infer d
1
< d
2
, d
2
< d
3
, d
n-1
< d
n

and so d
1
d
n
is a linear order under < and hence a chain.

Right to Left: Suppose that d
1
d
n
is a chain. It follows that for every d
i
e d
1
d
n
, d
i+1
= s(d
i
) and
hence d
i+1
e S
di
and so the following PC-Chain exists: d
2
e S
d1
, d
3
e S
d2
, d
n
e S
dn-1
.
T73: for any d
1
, d
n
, such that d
1
e h(w
@
) and d
n
e
TW
, d
1
makes d
n
possible if and only if d
1
<
d
n
.
Left to right: Suppose that d
1
is a truth-maker for the possibility that d
n
, then by the PC-postulate, there
exist some d
2
d
n-1
such that d
2
e S
d3
, d
3
e S
d2
, d
j
e S
dn-1
.

By Def. 87, for any d, d e
TW
, if d e S
d
,
then d < d. Hence, from d
2
e S
d1
, d
3
e S
d2
, d
n
e S
dn-1
we may infer d
1
< d
2
, d
2
< d
3
, d
n-1
< d
n
. By the
transitivity of < it follows that d
1
< d
n
.
Right to left: Supposing that d
1
< d
n
, where d
1
e w
@
and d
n
e
TW
, there must be some chain d
2
d
n-1

linking d
1
and d
n
and hence, by T72,it can be inferred that d
2
e S
d1
, d
3
e S
d2
, d
n
e S
dn-1
. Hence, by
PC-postulate, d
1
makes it the case that d
n
is possible.
Df99: For any d e
TW
, d is a leaf if and only if S
d
is empty.
A leaf is a world stage with no possible successor. While it is difficult to conceive how a world-stage
could be such that it is not followed by any other stage, I shall abstain from insisting that leaves are
impossible.
T74: For each d e
TW
there exists some h(w) e H such that d e h(w).
To say that each world-stage d belongs to a possible world is to say that each stage is a member of a
subset of
TW
that is a maximal linear order under < (i.e., that is a member of a maximal chain). Every
member of
TW
is either minimal or non-minimal. If it is minimal, i.e., if d = d, then d is a member of
every possible history (by T70). If d is not minimal, then, by back-ward linearity, there is a linear order of
stages S consisting each d such that d < d and the union of S and d is also a linear order. Now every d is
either a leaf or it is not a leaf. If it is a leaf, then it represents the end of the world and hence is part of a
maximal linear <-order of which it is the maximal member. Now suppose it is not a leaf, in which case it
has at least one successor. By T72, any sequence of possible successors is a chain. The union of all d
ancestors with any such successor chain will itself be a chain, and, indeed a maximal chain.
T75: For each d e
TW
there exists some d e h(w
@
) such that d e
TWd.

159

By T74, for each d e
TW
there exists some h(w) e H such that d e h(w). By T71, for every h(w) and
h(w) H, there exists some d e
TW
such that h(w)
d h(w), in particular, there exists some d such
that h(w)
d h(w
@
), where h(w) has been fixed as the world including d. But this means that there is a
member of h(w
@
), namely d that is a member of a world containing d (i.e., either d < d, d < d, or d =
d). Hence, d e
TWd unless d < d. But in that case, any d such that d < d will also be a member of
h(w
@
) (by downward closure) and hence it will be the case that d e h(w
@
) and d e
TWd.
T76: For every non-minimal d e
TW
there is some d e h(w
@
) such that d makes it the case that
d is possible.
By T73, for any d, d, e
TW
, d grounds the possibility of d just in case d < d. By T75, for each d e

TW
there exists some d e h(w
@
), such that d e
TWd
. By the latter, either d = d or d < d. If d < d, it
follows that d makes it the case that d is possible. If d = d then, given that d is non-minimal, it follows
that there is some d e h(w
@
)

such that d < d, and hence some d e h(w
@
)
,
meaning that d makes it the
case that d is possible.
With this machinery in place, it can be precisely shown how the modal (i.e., dispositional)
features of the actual world can make possible a plurality of ersatz worlds. Suppose that d e h(w
@
), d e
S
d
, and d h(w
@
). We know that the possibility of d is immediately grounded in the modal features of d.
Since d is a wholly determinate representation of how some moment might have been, its qualitative
character will determine the character of its set of possible successors S
d
(i.e., by A2 we are justified in
assuming that I determines J , since they should be meta factually equivalent).

Each d e S
d
is
indirectly made possible by d because, at d, there is some objective probability that each such d will
obtain. By continued iteration, d will ground the possibility of every d such that d < d in virtue of
making it the case that d has some objective probability of eventually obtaining.
128

Accordingly, any interpretation of a probabilistic branching-stage model with the PC-Postulate
that assigns true objective probabilities to all moments of h(w
@
) will include all and only those worlds
which are grounded by the modal features of the actual world. The I and J assignments that correctly
characterize each d e h(w
@
) force the values assigned by I and J with respect to merely possible world-
stages. If all of the modal features of the world are physical, then the resulting set of possible worlds will
be precisely the set of wide-nomologically possible worlds. Accordingly, there will always be a fact of the
matter as to whether any given ersatz world is also a metaphysically possible world, independently of our

128
See Mller (2011) for certain difficulties that arise when combining probability spaces defined on branching-time
structures and for a general strategy for overcoming these difficulties. Also, should there be a root, d, on
TW
,
every non-minimal member of
TW
would be assigned an absolute chance of obtaining relative to d.

160

knowledge of the modal facts (i.e., independently of our knowledge of the dispositional state of the actual
world at any given time).
The current account of metaphysical possibility supports S5 modality. We may now define
possibility and necessity as follows:
[V] V
M
(, w) =1 iff for some d e h(w
@
) there is some d such that d d and V
M
(, d) = 1
[V.] V
M
(., w) = 1 iff for each d e w
@
, and each d such that d d, V
M
(, d) = 1
[V] says that is possible just in case there is some stage, or moment, d of the history of the actual
world such that either is true at d or there is some possible future of d, of d, at which is true. (d will
evidently belong to at least one possible history).
[V.] says that is necessary just in case, for every stage d of the actual world, is true both at d and
every possible future stage d, of d.
By T75-76 [V.] effectively tells us that is necessary just in case it is true at every d e
TW
and,
conversely, V tells us that is possible just in case it is true at some d e
TW
. Accordingly, the
semantic structure of these clauses, is isomorphic to the standard possible world semantics, where to be
possible is to be true at some possible world and to be necessary is to be true at every possible world.
When there are no restrictions on which indexical items (worlds or stages) can be evaluated from any
other item, the resulting model is complete for the axioms of S5 modality.
It will be noticed that the truthmakers of modal claim are effectively construed in terms of
stages of the actual world, where each stage is a complete description of the actual world at a given time.
It should also be noted that world-stages are not the most fine-grained truthmakers. Typically it would be
expected that a genuine truthmaker of , will be an event that is properly included within a world-
stage. Upon reflection, however, this is not so obviously the case. Part of the difficulty of finding more
find-grained truth-makers concerns our ignorance of the extent to which the world might be causally
holistic. For instance, if gravitational force is spatially unrestricted, as it is represented as being, it
becomes difficult to see how one region of the universe could be absolutely causally isolated from any
other region. Ordinarily, we bracket out remote influences, knowing that they will rarely make any
difference to the final outcome. When taking up the view-from-nowhere, however, we cannot permit
ourselves such contextual or pragmatic luxuries. At any rate, the primary purpose of the diverging world
semantics is simply to show that possible worlds can be verified by the complete state of the actual
world.


161

2.3: Divergent Histories and the Denail of Presentism
It is widely thought that actualists run into trouble because ersatz worlds do not seem to be good
truthmakers for modal truths. This is a complaint raised by Lewis (1986a) among others. I conceded that
ersatz worlds cannot play to role of modal truthmakers but I denied that this is problem for actualists. In
order to justify this denial I have indicated how one can recover a large and varied set of ersatz possible
worlds on the basis of this-worldly modal truthmakers, namely actually instantiated probabilistic
dispositions (i.e., propensities). It turns out, however, that many of these truthmakers are states that take
place in the past. Moreover, it does not seem that we can find truthmakers in the present for all past and
future events. First, assuming that the world is not deterministic, it would be impossible for facts about
the present to be truthmakers for future events. Now most presentists believe that the future is open and
so this does not pose a problem. Can the presentist, however, provide truth-makers for all past events? It
may be the case that the present contains justifiers for all past events; that is, it might be the case that
given all present events and the laws of physics (extended to wide nomological possibility for the very
early universe), there is only a single possible past of the actual world. This would be the case if
information about all past states of the universe is conserved. Justification, however, is a matter of
knowledge while truth-making is a matter of truths. Knowledge is justified by evidence while truths are
verified by facts, or the way the world is. Now evidence is typically in the form of facts, but evidence for
believing need not assist in making the world such that is the case. First, there can be evidence for
even if is not true. Even where is true, justifiers can be distinct from truthmakers in the case of
testimony. Indeed, present events may justify true beliefs about the past precisely because they offer
testimony about those events. What we have is the testimony, but the testimony necessarily comes after
the fact. Claims about the past are not made true by present testimony; rather, the testimony is accurate in
virtue of some relation it bears to past events (presumably a causal relation and perhaps some form of
analogue resemblance).
The present account of HCA, then, would seem to be incompatible with presentism. In the
absence of presentism there would seem to be three possible contenders: eternalism, the growing block
theory of time, and the moving spotlight theory. On the growing block view the past fully exists but the
future does not exist and remains open in that there is no matter of fact as to what may or may not
happen in future times. On the moving spotlight view, the entire eternalist block universe exists, but there
is a sequence of prevledged moments, characterizing the passage of the present.
Accordingly, I observe a disanalogy between actualism and presentism. Then the actualist can
reduce possible world talk to actual world talk in the way I have indicated, provided that the past exists.
But I cannot find an analogous case for reducing past-time talk to present-time talk. Ersatz world talk can

162

be reduced to past stages of the actual world only if we are not talking about ersatz stages at ersatz times.
Ersatz items, being artificial and abstract, are not suitable candidates for the role of truthmakers, at least
not for matters that concern truths regarding concrete reality. Impossible ersatz worlds are just as real, qua
abstract object, as possible ersatz worlds, and hence the latter cannot ground anything that cannot just as
easily be grounded by the former. Likewise, false ersatz pasts are just as real as true ersatz pasts and so
the latter cannot ground anything that cannot be grounded by the former. If we permit the true ones to
ground possibilities it seems we must let the false ones do so as well, but that would be quite ridiculous
and contrary to HCA. The present account of modality, then, requires commitment to some alternative to
presentism.
3. Individual Hardcore Actualism (A4, A5)
This section presents two axioms pertaining to the modal natures of individual substances. The first, A4,
links the possible natural properties of an individual with manifestations of its dispositional ascriptions
while the second, A5, links these manifestations to the said dispositional ascriptions.
A4: Possible Natural Properties Instantiations are Possible Manifestations
!X!x!t
1
((X ^ Xxt
1
) !t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ (Xxt
1
m))))
If x possibly possesses some contingent natural property X it is only because x possesses a dispositional
predicate (typically iterated), the possibility of whose manifestation is factually equivalent with the
possible possession of said property. Xxt m should be understood to mean that the facts that
makes Xxt is possible are the same facts that make m possible. This will be satisfied even in the case
where X may be a necessary property, since, for each time t, the holding of X will still count as a
manifestation of holding X during the previous moment.
A5: Possible Manifestations are Partially Grounded by Instantiations of (Iterated)
Dispositional Ascriptions:
!D
(ccm)
!x!t(D
(ccm)
xt (m D
(ccm)
xt))
Dispositions only partially contribute to the possibility of manifestations because the circumstances under
which the manifestation fires must also be possible.
T77: (X ^ Xxt) .(X ^ Xxt)
1. X .X T18
2. Xx .Xxt S5
3. (X ^ Xxt) (.X ^ .Xxt) 1, 2. PL

163

4. (.X ^ .Xxt) .(X ^ Xxt) K Theorem
5. (X ^ Xxt) .(X ^ Xxt) 3-4. PL
T78: !X!x((X ^ Xxt) .D
(ccm)
(D
(ccm)
xt ^ (Xxt m)))
1. (X ^ Xxt
1
) !t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ (Xxt
1
m))) A4
2. (X ^ Xxt) .(X ^ Xx) T77
3. .(X ^ Xxt
1
) .!t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ (Xxt
1
m))) 1. K
4. (X ^ Xxt) .!t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ (Xxt
1
m))) 2-3. PL
5. !x!X!t((X ^ Xxt) .!t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ (Xxt
1
m)))) 4. UG
If X is a natural property and x is possibly X at t
1
then it is necessarily the case that there is some
dispositional predicate instantiated by x , at some t
2
, prior to t
1
, the possibility of the manifestation of
which is factually equivalent to the fact that x is possibly X.
T79: (Xxt m) (Xxt .m)
1. (Xxt m) (Xxt ^ m) F9
2. (Xxt ^ m) .(Xxt ^ m) S5
3. .(Xxt ^ m) (.Xxt ^ .m) K
4. (.Xxt ^ .m) (Xxt .m) PL, T
5. (Xxt m) (Xxt .m)
T80: .!x!X!t
1
((X ^ Xxt) .!t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ .m))))
1. (X ^ Xxt) .!t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ (Xxt
1
m))) T78
2. (Xxt
1
m) (Xxt
1
.m) T79
3. (X ^ Xxt) .!t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ (Xxt
1
.m))) 1-2. PL
4. (X ^ Xxt) .!t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ .m))) 3. PL
5.!x!X!t
1
((X ^ Xxt) .!t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ .m)))) 4. UG

T81: .!D
(ccm)
!t!x(D
(ccm)
xt X(X ^ (m Xxt)))
1. D
(ccm)
xt (m D
(ccm)
xt) A5
2. D
(ccm)
xt X(X ^ (D
(ccm)
x Xxt)) T55
3. D
(ccm)
xt X(X ^ (D
(ccm)
xt Xxt)) T58
4. D
(ccm)
xt (X(X ^ (D
(ccm)
x Xxt)) ^ (m D
(ccm)
xt)) 1, 3. PL
5. D
(ccm)
xt X((X ^ (D
(ccm)
x Xxt)) ^ (m D
(ccm)
xt)) 4. PL. CQT
6. (D
(ccm)
xt Xxt) ((m D
(ccm)
xt) (m Yxt)) F10
7. ((D
(ccm)
xt Xxt) ^ (m D
(ccm)
xt)) (m Yxt) 6. PL

164

8. X((X ^ (D
(ccm)
xt Xxt)) ^ (m D
(ccm)
xt)) Y(Y ^ (m Yxt)) 7. PL, CQT
9. D
(ccm)
xt Y(Y ^ (m Yxt)) 5, 8. PL
10. .!D
(ccm)
!x!t(D
(ccm)
xt Y(Y ^ (m Yxt))) 9. UG, NEC
Possible manifestations are factually contributed by natural properties all of which are, of course, also
dispositional properties.
T82: !x!X!t
1
!t
2
((X ^ (Xxt
1
^ t
2
< t
1
)) Y(Y ^ (Xxt
1
Yxt
2
)))
1. (X ^ Xxt
1
) !t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ (Xxt
1
m))) A4
2. (X ^ (Xxt
1
^ t
2
< t
1
)) D
(ccm)
(D
(ccm)
xt
2
^ (Xxt
1
m)) 1. CQT, PL
3. D
(ccm)
xt
2
Y(Y ^ (m Yxt
2
)) T.81
4. (D
(ccm)
xt
2
^ (Xxt
1
m)) Y(Y ^ (m Yxt
2
)) 3. PL
5. D
(ccm)
(D
(ccm)
xt
2
^ (Xxt
1
m)) D
(ccm)
Y(Y ^ (m Yxt
2
)) 4. CQT
6. (Xxt
1
m) ((m Yxt
2
) (Xxt
1
Yxt
2
)) F10
7. ((Xxt
1
m) ^ (m Yxt
2
)) (Xxt
1
Yxt
2
) 6. PL
8. ((Xxt
1
m) ^ Y(Y ^ (m Yxt
2
))) Y(Y ^ (Xxt
1
Yxt
2
)) 7. PL, CQT
9. (D
(ccm)
xt
2
^ (Xxt
1
m)) ((Xxt
1
m) ^ Y(Y ^ (m Yxt
2
))) 4. PL
10. (D
(ccm)
xt
2
^ (Xxt
1
m)) Y(Y ^ (Xxt
1
Yxt
2
)) 8-9. PL
11. D
(ccm)
(D
(ccm)
xt
2
^ (Xxt
1
m)) Y(Y ^ (Xxt
1
Yxt
2
)) 10. CQT
12. (X ^ (Xxt
1
^ t
2
< t
1
)) Y(Y ^ (Xxt
1
Yxt
2
)) 2, 11. PL
13. !x!X!t
1
!t
2
((X ^ (Xxt
1
^ t
2
< t
1
)) Y(Y ^ (Xxt
1
Yxt
2
))) 12. UG
For every natural property X that an object x may possibly instantiate, there exists some natural property
Y, and x being Y at t
2
partially grounds the fact that x is possibly X at t
1
, where t
2
< t
1
. That is, a natural
property can only be potentially possessed on the basis of actually possessing some (other) natural
property.
T83: .!X!x!t((X ^ Xxt
1
) .t
2
Y(Y ^ (Xxt
1
Yxt
2
) ^ t
2
< t
1
)))
1. (X ^ (Xxt
1
^ t
2
< t
1
)) Y(Y ^ (Xxt
1
Yxt
2
)) T82
2. .(X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ (Xxt
1
Yxt
2
)) 1. K
3. (X ^ (Xxt
1
^ t
2
< t
1
)) .(X ^ (Xxt
1
^ t
2
< t
1
)) T77
4. (X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ (Xxt
1
Yxt
2
)) 2-3. PL
5. .!X!x!t
1
!t
2
(X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ (Xxt
1
Yxt
2
))) 4. UG, NEC
Possible Natural Properties are partially grounded in actually possessed natural properties.
T84: .!X!x!t
1
((X ^ Xxt
1
) .t
2
Y(Y ^ Yxt
2
))

165

1. (X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ (Xxt
1
Yxt
2
)) T83
2. .Y(Y ^ (Xxt
1
Yxt
2
)) .Y(Y ^ (Xxt
1
^ Yxt
2
)) PG3
3. .Y(Y ^ (Xxt
1
^ Yxt
2
)) .Y(Y ^ Yxt
2
) CQT
4. (X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ Yxt
2
) 2-3. PL
5. !X!x!t
1
!t
2
((X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ Yxt
2
)) 4. UG
Anything that possibly possesses a natural property necessarily possesses some natural property.
T85: .!x!X!t((X ^ Xxt
1
) .t
2
Y((Y ^ Yxt
2
) ^ t
2
< t
1)
)
1. (X ^ (Xxt
1
^ t
2
< t
1
)) (X ^ (Xxt
1
^ t
2
< t
1
)) T82
2. ((X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ Yxt
2
)) T84
3. (X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ Yxt
2
)) 1-2. PL
4. .!X!x!t
1
!t
2
((X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ Yxt
2
)) 3. UG, NEC
For any time that an object possesses a natural property there is some earlier time at which it also
possesses some natural property.
T86: Necessity of Concreteness (NC): .!x!t
1
!t
2
((Cxt
1
^ t
2
< t
1
) .Y(Y ^ Yxt
2
)))
1. (X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ Yxt
2
)) T85
2. Cxt
1
X(X ^ Xxt
1
) Def. C
3. Cxt
1
X(X ^ Xxt
1
) 2. PL
4. (Cxt
1
^ t
2
< t
1
) X((X ^ (Xxt
1
^ t
2
< t
1
)) 3. PL, CQT
5. X((X ^ (Xxt
1
^ t
2
< t
1
)) .Y(Y ^ Yxt
2
) 1. CQT
6. (Cxt
1
^ t
2
< t
1
) .Y(Y ^ Yxt
2
) 4-5. PL
7. (Cxt
1
^ t
2
< t
1
) .Y(Y ^ Yxt
2
) 6. DR5
8. (Cxt
1
^ t
2
< t
1
) .Y(Y ^ Yxt
2
)) K.
9. .Y(Y ^ Yxt
2
)) .X(X ^ Xxt
2
) CQT
10. .X(X ^ Xxt
2
)) X(X ^ Xxt
2
) T
11. X(X ^ Xxt
2
) Cxt
2
2. PL
12. .X(X ^ Xxt
2
) .Cxt
2
11. K
13. (Cxt
1
^ t
2
< t
1
) .Cxt
2
8- 9, 12. PL
14. .!x!t
1
t
2
((Cxt
1
^ t
2
< t
1
) .Cxt
2
) 13. CQT, NEC
Corollary: If is the cardinality of concrete objects, then the cardinality of concrete objects is necessarily
.

166

For any object x, at any time t
1
at which x can possibly be concrete, x is necessarily concrete at every time
t
2
such that t
2
is earlier than t
1
. This shows two things. First, every object that can possibly be concrete is
concrete in every possible world at some time or other. Second, it is not possible for something to become
concrete, which means that everything is concrete now, or any time in the past, has always been concrete.
It has not, however, been demonstrated that nothing can cease to be concrete while continuing to
exist. NC does, however, show that something could only cease to be concrete if it ceases to be concrete
at the same time t in every possible world. If there is any world-time pair w, t at which x is concrete,
then, in every world x is concrete at every time prior to t. Assuming a reasonable amount of
indeterminism in the world (and I must in order to back HCA), it may generally be assumed that no object
is such that is necessarily ceases to be concrete at some one specific time unless we have some definite
reason for thinking otherwise.
Do31: Existence Monism is the doctrine that there exists exactly one concrete object.
x(Cx ^ !y(Cy y = x))
T87: The truth or falsehood of Existence Monism is a necessary matter.
This follows immediately from the corollary of T100.
T88: Necessity of Abstractness (NA):
.!x(Axt .Axt)
1. (Cxt
1
^ t
2
< t
1
) .Cxt
2
T100
2. Axt
1
Cxt
1
Def. A
3. (Axt
1
^ t
2
< t
1
) .Axt
2
1,2 PL
4. .Axt
2
(Axt
1
^ t
2
< t
1
) 3. PL
5. Axt
2
(Axt
1
^ t
2
< t
1
) 4. Def.
6. Axt
2
(Axt
1
\ t
2
< t
1
) 5. PL
7. Axt
2
(.Axt
1
\ t
2
< t
1
) 6. Def.
8. Axt
2
(t
2
< t
1
\ .Axt
1
) 7. PL
9. Axt
2
(t
2
< t
1
.Axt
1
) 8. PL
10. (Axt
2
^ t
2
< t
1
)

.Axt
1
9. PL
11. .!x!t
1
!t
2
((Axt
2
^ t
2
< t
1
)

.Axt
1
) 10. UG, NEC
As is the case of NC, NA establishes two things. First, anything that could be abstract (i.e., non-concrete)
is non-concrete in the actual world at some time or other. Second, for any object x and time t
2
, if x is

167

abstract at t
2
in any possible world, then, for every time t
1
that is later than t
2
, x is necessarily abstract at t
1
.
That is, nothing can ever cease to be abstract.
By NC it is impossible for anything for anything to become concrete and by NA it is not possible
for anything to cease to be abstract. This is, of course, because abstract objects, having no powers, have
no potential of becoming concrete. Something could only become concrete if something else ceased to be
abstract. However, the possibility of something ceasing to be concrete, or becoming abstract, has not been
completely eliminated.



















Copyright Adam Labecki 2013

168

CHAPTER NINE:
THE NECESSITY OF FUNDAMENTAL PROPERTIES

Objects are what is unalterable and subsistent;
their configuration is what is changing and unstable.
Wittgenstein, Tractatus Logico-Philosophicus 2.0271

The present chapter shall be committed to (1) motivating the claim that all strongly intrinsic natural
properties, and hence all fundamental properties, are necessarily possessed by their bearers and (2)
considering the implications of this claim. I hope to motivate the counterintuitive claim that the essence of
a fundamental object is precisely the set of all of its fundamental properties. The case shall be made in
three phases. First, it will be argued that we are generally quite ignorant of the intrinsic properties of
observable objects, much less do we possess knowledge of the strongly intrinsic properties of objects.
Some work to this end has already been accomplished from the cosmological considerations that were
pertinent to wide nomological possibility. The purpose of this first step is to undermine the apparent
evidence that the intrinsic properties of objects undergo change. Second, the well-known Problem of
Temporary Intrinsics (PTI), and the related Problem of Contingent Intrinsics (PCI), suggests that intrinsic
properties are both temporally and modally rigid. This line of argument will prove wanting in that it
cannot be rigorously formalized in any way that is not clearly circular. Nevertheless, two new arguments
for the fixity of strongly intrinsic properties are introduced. Third, of the three standard ways of
responding to PTI (or rather, the problem of temporary strong intrinsics), the relational approach will
be the strongest candidate since the other two options have already been eliminated. A series of theorems
follow regarding the necessity of fundamental properties. All contingent properties will turn out to be
explained, in part, by natural relations obtaining between fundamental objects, which means, by A1, that
priority monism and existence monism must be denied.
1. The Case against Observable Intrinsic Properties
There are a multitude of reasons to suppose that we are generally ignorant of the intrinsic properties of
observable, physical, objects. Indeed, many philosophers have already noted that the properties
discernible by physics do not appear to be intrinsic:
If we look at the properties of physical objects that physicists are prepared to allow them such as
mass, electronic charge, or momentum, these show a distressing tendency to dissolve into relations
that one object has to another. What, then, are the things that have these relations to each other?
Must they not have a non-relational nature if they are to sustain relations? But what is this nature?
(Armstrong 1968: 74-75).

169

I think we should acknowledge as a possible, interesting, position one we might call Kantian
physicalism. It holds that a large part (possibly all) of the intrinsic nature of our world is
irretrievably beyond our reach, but that all the nature we know about supervenes on the causal
cum relational nature that the physical sciences tell us about (Jackson 1998: 24).
We have argued that entanglement as described in QM teaches us that Humean supervenience is
false, and that all the properties of fundamental physics seem to be extrinsic to individual objects
(Ladyman and Ross 2007: 151).
No known physical properties are self-sufficient, for all are causally dependent on ambient
conditions (R. Harris 2010: 106).
129

Similar conclusions, arising on the basis of various considerations, are also made by Adams (1997),
Lewis (2009), and Langton and Robichaud (2010). While such observations are not uncommon and
difficult to contest, metaphysicians are typically forgetful of them when considering particular theories
and in presenting counter-examples. For instance, it is typical to regard comparably homely properties
like shape as intrinsic despite the fact that even mass and spin appear to be extrinsic.
Arguments to the effect that we are ignorant of the intrinsic properties of things seem to come in
two basic varieties: arguments from physics and epistemically arguments, where the latter are typically
Kantian in nature. I shall focuses on considerations from physics, since, in my own judgment they are
more persuasive and less controversial.
At least three aspects of contemporary physics suggest that observational properties are not
intrinsic. The first two are similar, however, in that both give rise to arguments from what Harris (above)
called ambient conditions. In particular, they are arguments from ubiquitous ambient conditions
namely global physical fields that have tremendous effects on the resulting phenomena. Bauer (2011), for
instance, argues that mass is an extrinsic property as follows:
(1) The property inertial mass of any fundamental particle, a, is generated by as immersion in
the Higgs field.
130

This is merely a facet of the Standard Model of particle physics and explains why physicists have been so
eager to observe the Higgs boson. In effect, the interaction between a particle a and the Higgs field, which
is a universal scalar field, converts some of the energy of the field into the mass of the particle. Indeed,
even the mass of the Higgs boson is believed to be obtained by this interaction. Since the Higgs field has
this capacity, it must, by A2, be in virtue of some natural property that it does so. Likewise, the particles
must have properties that dispose them to acquire mass when immersed in the Higgs field, which

129
The first two passages were cited in Langton and Robichaud (2010), who also endorse the view that we cannot
know the intrinsic properties of things.
130
I modify the premise slightly by replacing disposition with property. Bauer himself notes that the argument
goes through in the same way if one regards such properties as categorical. I, of course, regard them as both
dispositional and categorical.

170

properties are not mass properties but dispose their bearers to acquire mass under certain conditions. The
magnitude of the acquired mass is proportional to the strength of the interaction; i.e., different particles
have different capacities to absorb energy from the Higgs field.
131

(2) The nature of the relation between a fundamental particle a, and the Higgs field meets the
condition for extrinsic grounding.
According to the Standard Model, particles obtain their mass in virtue of sustained immersed in
the Higgs field (i.e., were the Higgs to suddenly disappear, the other particles would lose all inertial
mass), Moreover, the Higgs field is distinct from the particles with which it interacts. It matters not if a
particle and the area of the Higgs field with which it interacts happen to be co-located (although, again,
particles may typically lack determinate locations). Supersubstantialism has already been rejected and so
one cannot try to attempt to identify a particle with the part of the Higgs field with which it interacts.
Hence, the inertial mass of a particle is partly explained by some relation it bears to a distinct object,
which is to say:
(3) The inertial mass of fundamental particles is extrinsically grounded (from 1 and
2).
132

It should also be noted that the interaction between a particle a and the Higgs field does not result
in a having a determinate mass, it merely grounds the fact that a has some mass.
133
The specific mass of a
particle also depends upon other objects in its environment which determine its freedom of motion (and
hence the rate at which it might be accelerated by a given force). Given that natural properties are
determinate properties, mass turns out to be doubly extrinsic, which gives some reason to suppose that it
cannot even be nearly fundamental.
The above argument specifically concerns inertial mass, which is at least conceptually distinct
from both passive gravitational mass and active gravitational mass. GR assumes that passive gravitational
mass is physically identical with inertial mass. Whether or not inertial mass and active gravitational mass
are physically identical is a somewhat open question but the inertial mass of an object is never observed

131
The mechanism of the absorption is, somewhat metaphorically, explained in terms of friction, some particles
drag onto the Higgs field more strongly than others, thereby acquiring more mass.
132
Whether we should say that the resulting mass is factually equivalent to its interaction with the Higgs field or
caused by this interaction is somewhat tricky, since the having of the mass seems to be simultaneous with said
interaction. I would leave the matter open for the moment. If it is to be construed as broadly causal (and hence a
matter of grounding), then it would be said to emerge from the interaction (this would be a relatively non-exotic
form of emergence).
133
Difference in strength of the interaction with the Higgs field is one factor contributing to the magnitude of the
actual mass of a particle, but it does not determine the mass by itself.

171

to differ from its active gravitational mass.
134
The Einstein Equivalence Principle requires that inertial
mass and passive gravitational mass be equal. The Law of Conservation of Momentum requires that the
active and passive gravitational mass of any object should be equal.
135
By transitivity of equality, the
inertial mass of an object must be equal to it its active mass. Systematic coincidence of magnitude does
not strictly entail identity, but it does provide strong evidence for it. At any rate, even if these forms are
distinct in reality, it is probably nomologically impossible for an object with no inertial mass to have any
mass at all. Mass, of course, is only one property, but within the Standard Model it is a very central
property and is closely related to one of the four fundamental forces, namely gravity.
A second argument from physics comes from cosmology and concerns just the cosmological
features discussed in Chapter Eight.
136
It was observed that the separation of the four fundamental
forces is dependent upon the presence of a global minima or cosmological constant. The present state of
low energy physics is not merely an effect of past events, since the continued presence of this global field
(or fields) is required if the four fundamental forces are to exist as they do.
Since most of the basic properties of basic particles directly concern their roles as force
carriers, none of these properties can be regarded as intrinsic. Moreover, each of the basic observational
properties that a fundamental particle or field may have is understood in terms of how it interacts with one
or more of the fundamental forces. Sufficient change in the global minima or cosmological constant
results in a complete change in which sort of properties a physical object may manifest.
Common to these first two arguments from physics, is the fact that all observational properties
are partially explained by the omnipresence of one or more one or more global fields (the Higgs field and
the cosmological constant). These would be the most universal examples of the sort of ambient
conditions mentioned in the quotation from R. Harris above. When we consider whether or not a
property is intrinsic, we are not free to ignore non-salient features of the world. Saliency is a relative
and conventional concept that is out of place when we take up the view from nowhere. That is, from an
objective point of view, every actual and positive condition whose absence would have resulted in a
distinct outcome must be though to contribte to said outcome.

134
The Einstein Equivalence Principle postulates that inertial mass and passive gravitational mass must be equal.
The Law of Conservation of Momentum requires that active and passive gravitational mass be equal. By transitivity
of equality General Relativity requires that the inertial mass of an object must be equal to it its active mass.
135
These are absolute values, so if there is such a thing a negative mass, the two masses would not be strictly the
same.
136
To my knowledge this argument from cosmology has not been previously observed.

172

There are additional arguments from physics that seem to show that at least most observational
properties are extrinsic. Perhaps the most common variety arises from quantum entanglement.
137
When
two or more particles are in an entangled state relative to some determinable property X, there is,
according to standard interpretations of quantum mechanics,
138
no fact of the matter as to the whether or
not each particle in the system has some determinate value for X. The system as a whole, however, can be
assigned a determinate value for X, although it is not clear how this value should be physically interpreted
(i.e. it does not correspond to a possible observation).
139
The best known cases of entanglement regard the
position of the particles in question. Spatial position is not generally considered to be an intrinsic
property, so such cases do not of themselves threaten the intrinsicality of physical properties.
140

In addition to position and velocity, the spin and charge of particles are prone to entanglement.
Only the system as a whole has a determinate value for these properties. Accordingly, it would appear
that the observable properties of the fundamental particles are typically extrinsic.
A. Ney (2010) argues that the argument from entanglement does not completely dispel intrinsic
properties from physical objects. She notes that one can save intrinsic physical properties by adopting a
strong form of wavefunction realism
141
which leads to a strong form of configuration-space realism.
142

Her argument runs as follows:
1. There exist entangled states.

137
See Teller (1986, 1990), Maudlin (2007), Esfeld (2004), Ladyman and Ross (2007), Morganti (2009), Darby
(2012). There are reasons to believe that entanglement is quite a common phenomenon, especially given the fact that
the Schrdinger equation, which determines the evolution of the wavefunction, has it that systems of particles tend
to become more entangled over time and it never predicts collapse from a superposition to a determinate position
138
According to the standard interpretation of quantum mechanics the wavefunction, , a system that is entangled
with respect to a property X is understood to include all physical information about the system relative to property
X. It is a consequence of Bells Theorem that if there is some hidden variable such a variable could not be a local
part of the system. The mechanics proposed by Bohme introduce such non-local variables, the resulting physics
turns out to be extremely holistic, and hence on that view physical properties will also (at least typically) be
extrinsic.
139
This is especially the case, because the value is often zero.
140
The spatial positions of a plurality of particles xs should, however, contribute to the intrinsic properties of any
object composed of the xs.
141
Realism in particle physics practically requires that one takes the wavefunction with some ontological
seriousness. If the wavefunction provides complete physical information regarding the state of a system and if it is
regarded as a mere probability equation, then one will have difficulty finding anything in QM that one can call
physical.
142
The configuration space, also known as 3N-space, of an entangled quantum system has three dimensions for each
particle in the system. If entanglement is universal as suggested by Schaffer (2010) then the dimensionality of the
universe would be three times as great as the cardinality of fundamental particles (or field excitations) in the
universe.

173

2. These states cannot be adequately characterized as states of something inhabiting a three-
dimensional space, but rather must be characterized as spread out in a higher-dimension
configuration space.
3. So, (from 1 and 2) there is something of the kind that must be characterized as spread out in a
higher-dimension: call this the wavefunction.
4. So, the wavefunction exists.
The argument can be resisted at premise (2), in that one can take the role of the wavefunction seriously
without regarding it as a physical object existing in configuration space. One might, for instance, interpret
it as indicating state of a system in wholly dispositional terms, which do not directly reveal the categorical
state of said system. However, given that Neys purpose is simply to show that QM permits us to speak of
intrinsic properties of some kind of physical object, the argument surely succeeds, since (2) is a perfectly
natural interpretation of the ontological implications of QM.
If one is willing to adopt such a strong form of wavefunction realism, then the Argument from
Entanglement can preserve intrinsic properties, although they will not belong to what are known the
fundamental particles and hence they will not be what metaphysicians typically mean when they speak
talk about intrinsic physical properties. Moreover, the middle-sized objects of human experience will also
lack intrinsic properties, since those properties are not described in terms of 3N configuration space.
Three-dimensional space and micro-particles are merely well-founded phenomena. Moreover, the view is
strongly holistic. The universal wavefunction is the one object whose intrinsic properties ground particles,
tables, persons etc.
143
I shall not go into the virtues and vices of this ontology at present, but will note that
even on Neys account, ordinary examples of intrinsic properties will not turn out to be intrinsic. There
will be intrinsic change, however, but not of the kind with which we are familiar and only one object
undergoes intrinsic change.
Perhaps the best known arguments to the effect that observational properties are not intrinsic are
epistemic in nature. In this context the work of Kant, especially as interpreted by Langton (1998),
immediately comes to mind. Langton emphasizes the notion of receptivity in Kants epistemology. On
this view, we cannot know things as they are in themselves, that is, as they are intrinsically, because we
can only know them relative to our forms of intuition (space and time) and the categories of the

143
Ney concedes to Schaffer (2010) that it is likely the case that the entire universe is one entangled quantum
system.

174

understanding. In this way observational properties turn out to be relational in that they are partially
explained relative to the cognitive capacities of finite knowers.
144

While it may be the case that part of the relational nature of observational properties arises from
the fact that, qua observable, they are extrinsically grounded in our cognitive capacities and measuring
devices, I have chosen to emphasize the ways in which the observational properties of objects are
extrinsically grounded in their relations to one another. I have done so for three reasons. First, this line of
argument will appeal even to direct realists or to those who would otherwise resist the Argument from
Receptivity.
145
Second, I believe that paying attention to the way in which observable physical properties
are extrinsic on purely physical considerations lends us greater insight into the nature of the phenomena.
Third, the Argument from Receptivity hits too close to home, as it were, due to the fact that I am more
or less applying the categories of the understanding to Reality in itself. I am, of course, doing so with
caution, but I am, naively or not, proceeding as if metaphysics were possible without turning to Kantian
idealism.
2. Problems for Temporary or Contingent Intrinsics?
Having argued that we have no strong evidence in favor of intrinsic change, I shall now proceed to
consider some arguments against the very possibility of intrinsic change. None of these arguments will be
entirely decisive. In particular, the best known argument against intrinsic change will turn out to be
circular. I shall propose two arguments that conclude that strongly intrinsic properties, and hence
fundamental properties, are necessary properties. Each such argument, however, requires at least one non-
trivial assumption. In each case the assumption in question is very plausible but, nevertheless, deniable.
2.1 On the Problem of Temporary Intrinsics
David Lewis (1986a) presents what has become a well-known case against there being temporary
or contingent intrinsic properties: it seems as though change in intrinsic properties easily leads to
contradiction. Suppose that Ruth is sitting at some time t and not-sitting at some time t. By supposition, it
follows that Ruth is both sitting and not-sitting. That is a contradiction. Let us call this the Problem of
Temporary Intrinsics (PTI). The problem also arises in the modal case. Suppose that Ruth is 5`7 at w and
Ruth is not 5`7, but 5`9 at w. It follows that Ruth is both 5`7 and not 5`7. Contradiction again! Let us call

144
This general form of argument, however, goes back to the Pyrrhonian Skepticism of Sextus Empiricus.
145
The Direct realist, of course will want to deny the entire reality versus appearance distinction. Those who
maintain a Representational Theory of Knowledge will have to directly wrestle with Kantian objections, but will try
to maintain that something like the primary-secondary quality distinction can be sustained (where the intrinsic-
extrinsic property distinction is the contemporary form of the older distinction).

175

this the Problem of Contingent Intrinsics (PCI). The Argument from Temporary Intrinsics is typically
spelled out as follows:
1. Ruth at t is sitting and Ruth at t is not sitting (Assumption for Reductio)
2. <Ruth at t> = <Ruth at t = Ruth> (By I1?)
3. Ruth is sitting and Ruth is not sitting. (1-2 I2?)
To obtain an argument for PCI, simply replace t, t with w, w.
It turns out that this version of PCI cannot be precisely formalized and, as far as I am aware, there is no
more precise account of the argument is to be found in the literature. The first line can be formulated as
follows:
1. Srt ^ Srt
Line (2), however, is stubbornly resistant to formalization. On the surface it seems to be saying something
like Ruth at t is identical with Ruth at t. But rt = rt which is non-grammatical; it is not formula of
the object language of this treatise nor, to my knowledge, is it a formula of any version of temporal logic.
Even the Permanence of Existence (PE) only tells us that if there is some time that x and y are identical
then they are always identical. One might suppose that the following is promising E!xt = E!xt but two
events at different times cannot be identical. Indeed such expressions are probably better served as
speaking about two distinct ersatz temporal parts of x. Semantically speaking, (2) is probably meant to say
that r refers to the same object at both t and t, but that seems to be all that it is saying. Moreover, even if
(2) is meaningful, (3) is not an analytic entailment from (1) and (2).
A tacit assumption in PCI is something like the following:
(2.1) !x(t
1
(x, t) (x))
146

I do not understand what (2.1) is supposed to mean. Once we introduce time parameters, all formulas will
have truth-vales which are relative to times, even formulas which are true all of the time. Is the
consequent supposed to be true of the present time? In that case (2.1) seems to be false: can be true at
some time without being true of the present. Likewise, if the consequent is intended to be true at all times,
it is also something that seems obviously false. Finally, if the consequent, like the antecedent, is only
supposed to be true at some time or other, then it will be clearly valid (it would be a tautology) but it will
be of no use in producing a reductio. Since there are no further alternatives, (2.1) cannot be used to save
PTI.

146
Typically the assumption is in terms of tense rather than time: !x(P(x) \ F(x)) (x))

176

(3) is also semantically ambiguous, in that we are not told when it is that Ruth is both sitting and
not-sitting. Again, if it is intended to be true all of the time or at the present time, then it does not follow
from the premises. Indeed, it does not even follow that there is any one time at which she is both sitting
and not sitting.
Of course, the inference from (1) and (2) to (3) is supposed to follow from I2. In order to evaluate
the inference we need to be clear as to how I2 is supposed to incorporate temporality. Just as there are
questions about how second-order comprehension axioms should incorporate modality, we must also
decide how the logic of identity should incorporate temporality. I have already posited the following,
which is surely the most natural temporalization of I2:
!t(x =
t
y ((x, t) (y, t)))
147

This construal of I2 will not get temporary intrinsics into trouble. The only precise way to present a
rigorous PTI argument on the basis of I2 is to render it as follows:
I2*: !t
1
t
2
(x =
t1
y ^ x =
t2
y) ((x, t
1
) (y, t
2
))
From (1) above in conjunction with x =
t1
y ^ x =
t2
y (as a replacement for (2)) we can surely derive a
contradiction on the basis of I2* but this is because I2* begs the question regarding the possibility of
temporal change.
148
One would only choose I2* over my proposal if one already had it out for temporary
intrinsics. Moreover, I2* will not only pose a problem for temporary intrinsics but temporary properties
altogether. Hence I2* is surely too strong, and even if it were restricted to intrinsic properties it would be
too heavy-handed to use as a basis for the rejection of temporary intrinsics. I tentatively conclude, then,
that there is no valid and non-circular PTI-style argument.
Perhaps there is another way of making Lewis problem with temporary intrinsics more precise.
For instance, he does not want to force intrinsic property instantiations to be relativized to times, since
that would seem to make them relative to something other than the object in question and hence non-
intrinsic. This is no doubt the motivation behind premise 2.1 above, but 2.1 is malformed and hence
impermissible. In Chapter Five I construed as intrinsic those properties whose possession is not partly
explained in terms of some relation that the possessor bears to some distinct object. Relations to times are

147
Given PE one can go on and say t
1
x =
t1
y !t
2
((x, t
2
) (y, t
2
)))
148
Hansson (2007) also argues that any way of construing PTI is ultimately question-begging, although his analysis
differs from mine.

177

exempt, since they are not objects. They are not, in and of themselves, substantial relations. Lewis, it
seems, has gotten himself too hung up on matters of surface grammar.
149

2.2 The Argument from Strong Nominalism
Inasmuch as Lewis is not merely motivated by his commitment to four-dimensionalism and inasmuch as
he has not been deceived by talk that is relative to times, it seems he means to say that intrinsic should
be construed as follows:
A6a*: Intrinsic X: X !x!t(Xxt (Xxt E!xt))
150

The instantiation of an intrinsic property X by an object x at a time t is factually equivalent to the
existence of x at t. While this formula is in the form of a definition is it too strong, in my judgment, to
simply introduce as a definition of intrinsic. The definition, does, however, carry some intuitive appeal.
Intuitively speaking, an intrinsic property is a property that is possessed simply in virtue of the possessor.
Indeed, one might say that the intrinsic properties of an object just are their bearer, descriptively
conceived. Alternatively, those who hold the bundle theory of objects simply identify objects with
bundles of properties. It is clearly the case that, construed in this way, intrinsic properties will be
necessary properties by way of F8 (this will be explicitly demonstrated momentarily).
Can more be done to motivate A6a*? While A6a* is not objectionably circular like I2*, it is quite
clear that, in conjunction with NE and PE, it will result in the fixity of intrinsic properties. It is an
improvement, I think, in that it brings the debate into the right focus. That is, I believe that PTI is an
argument in the sense that the first proof for the existence of God in Descartes Meditations is an
argument. In both cases the argument is either circular or invalid, but on a charitable reading one might
say the point is really a matter of understanding the very concepts at issue. Descartes wants to guide his
reader to the realization that to understand what is meant by God is to understand that something exists
that answers to the concept of God. Lewis, it seems, wants to say that to understand what is meant by
intrinsic is to understand that the intrinsic properties of an object must be both temporally and modally
invariant. A6a* is my way of trying to formalize this judgment.
I suspect, however, that many will remain unconvinced, and rightly so. A6* seems right but it
also seems wrong. Part of why it seems wrong is because we typically take ourselves to be familiar with
temporary intrinsic properties. I have addressed this matter above but we should still seek some positive

149
To a spacetime supersubstanialist like Lewis, relations to time will be substantial (i.e., they are relations to
genuine things), but PTI is supposed to lead one to supersubstantialism (via four-dimensionalism). Accordingly, one
would not want to presuppose supersubstantialism within a PTI argument.
150
Lewis does not, of course, define intrinsic in this way in any of his works but, in the terminology of the present
treatise this is how one would define intrinsic if it is to be a property that is just possessed simpliciter.

178

reasons for the acceptance of A6a*. After all, we know that there is a lot of change of some kind or other.
Perhaps we never directly observe changes in an objects intrinsic properties, but it is still natural to
suppose that the changes we do observe are typically the result of intrinsic changes.
I shall present an informal argument in favor of A6a*, beginning with a statement of the
following doctrine:
Strong Nominalism is the doctrine that there are no universals or tropes before things (ante
res) or in things (in rebus).
Strong nominalism so defined is compatible with conceptualism or post res universals and tropes. The
judgment lying behind strong nominalism is that when one attributes a property to an object one is not
referring to some relation between a thing signified by a predicate and the predicated object. There is just
the thing. To be sure, we want to speak of the way the thing is but the way of the thing is not something
distinct from the thing. In some respects this view is similar to the bundle theory, but the reduction goes
in the opposite direction. The strong nominalist does not drop the substance in favor of some set of tropes;
rather, he drops the set of tropes in favor of the substance. The substance is not conceived here as a bare
particular or substrate but simply as the thing itself. When you have the thing itself you also have the way
that it is. When we think and speak about the way something is we must make use of predicates.
151

Moreover, we will frequently apply the same predicate to many things. When both x and y are F, it is not
that there is a thing (i.e., a universal) that they have in common; nor does it mean that they each possess
numerically distinct but qualitatively similar things (i.e., two F tropes). If tropes can bear primitive
similarity with one another why cant we just do without the tropes and let objects bear primitive
similarities with one another? The strong nominalist, then, may well be inclined to adopt resemblance
nominalism. If so, properties might be construed as resemblance classes among objects. Perhaps the
strong nominalist will adopt some other solution to the problem of universals (although I am not sure
what) or perhaps he will decide to be an ostrich nominalist.
Why would one want to be a strong nominalist? Probably because one cannot believe that there
are such things as mind independent universals or tropes.
152
It is probably about as simple as that. I shall
forego the presentation of various intuition pumps intended to show how odd it is to suppose that there
are such entities. Those who are dedicated to universals and tropes are likely to bite the bullet over such

151
Perhaps it is possible to have some sort of non-propositional knowledge of certain objects, but whatever non-
propositional knowledge might be, it is not scientifically communicable. Any such insight into the nature of things
can, at most, express itself in guiding the way we develop our proposition-based theories.
152
This is effectively van Inwagens (2011) stance toward what he calls constituent ontologies by which he has in
mind both the bundle theory and substrate theory. His plea is that he does not understand what it means to say that
objects are either composed of tropes or a bare particular plus a set of universals.

179

cases anyway and I do not have any new reasons to give for why one should not include universals or
tropes into ones ontology. If, however, you are already suspicious of using universals or tropes to explain
similarities and difference between objects you will be on board at this point.
Now, if the way something is intrinsically is not in virtue of participation in universals or the
presence of tropes, in what is it in virtue of? This is where the argument makes an inductive inference: if x
is intrinsically F and there are no universals or tropes then x being F is nothing over and above x itself,
which is to say, x being F is factually equivalent to x exists. The breakdown of the argument is as
follows:
(1) If F is an intrinsic property and Fx is true, then Fx is factually equivalent to x
participates in universal F, or an F-trope is present in x (or is part of x), or x
existing.
(2) If the instantiation of an intrinsic property F is factually equivalent to the participation
in a universal F, then F is a mind-independent universal.
(3) If the instantiation of an intrinsic property F is factually equivalent to the presence or
inclusion of an F-trope, then the trope in question is mind-independent.
(4) There are no mind independent tropes or universals (Strong Nominalism); therefore,
(5) If F is an intrinsic property and Fx is true, then Fx is factually equivalent to x
existing.
Premise (1) presupposes that there are only three ways in which intrinsic property instantiations might be
explained. It is, however, missing compositional properties, properties that a composite has in virtue of its
parts and their arrangement. The argument as it stands could only apply to non-compositional intrinsic
properties, and hence strongly-intrinsic properties. When restricting ourselves to natural properties, the
result will be that all fundamental properties instances are factually equivalent to the existence of their
bearers. When restricted in this manner, the Argument from Strong Nominalism seems to be quite strong.
The first premise appears to be true when adding compositional properties to the disjunction. To be
sure, I could be missing some additional way a property might be intrinsic, but it would seem to be one
that everyone else is as well. I do not believe that anyone would want to deny the second or third
premises. The most disputable premise is the fourth, being the affirmation of Strong Nominalism.
Indeed, this inference is already suggested by the original definition of intrinsic: X !x(Xxt
Y
n
y
1
y
n-1
((Xxt Y
n
y
1
y
n-1
xt) (Py
1
xt...^...Py
n
xt)). If being X is not partially explained by some
relation of any proper parts of x (i.e., if X is non-compositional) is there any relation that might partially

180

account for its being X? It can only be even partially equivalent to some relation involving x alone (and t).
The most basic self-relation is identity, relative to which existence has been defined. If x being X is partly
explained by the existence of x, then what facts make up the remainder? The most straightforward
response, I think, is that in such instances the conditional is trivially satisfied by the fact that nothing
partially explains Xxt but, rather, that Xxt is factually equivalent to E!xt.
2.3 A Proof from Unique Modal Equivalence
. There is also a second way in which one might derive the necessity of fundamental properties.
Given A3 fundamental properties play an anchoring role in that other natural properties should
ultimately be manifestations of fundamental properties, typically involving iterated dispositional
ascriptions. Accordingly, it is reasonable to suppose that fundamental properties should be non-
redundant. There are various senses in which a property may be non-redundant. One way in which a
property may be said to be non-redundant is if it bestows some natural possibility upon its bearer which is
not bestowed by any other property. Let us call this claim Unique Modal Equivalence (UME).
A6b:* Unique Modal Equivalence (UME): (X ^ t
1
< t
2
) Y((Y ^ .(Xxt
1
Yxt
2
))
For each fundamental property X and times t
1
and t
2
, where t
1
< t
2
, there is some natural property Y such
that, necessarily, something is X at t
1
if and only if it is possibly Y at t
2
. This is effectively to say that
each fundamental property X predisposes its bearer to possess some natural property Y that that no non-X
object is predisposed to possess. Given A6b*, it can be shown that all fundamental properties are
necessary properties.
T:* Necessity of Fundamental Properties:
.!X!x!t
1
t
2
((X ^ Xxt
1
) .Xxt
1
)
1. (X ^ t
1
< t
2
) Y(Y ^ .(Xxt
1
Yxt
2
)) UME
2. (X ^ t
1
< t
2
) Y.(Xxt
1
Yxt
2
) 1. CQT, PL
3. (X ^ t
1
< t
2
) Y.(Xxt
1
Yxt
2
) 2. PL
4. (X ^ t
1
< t
2
) Y(.Xxt
1
.Yxt
2
) 3. K.
5. (X ^ t
1
< t
2
) Y(.Xxt
1
Yxt
2
) 4. S5
6. (X ^ t
1
< t
2
) (.Xxt
1
YYxt
2
) 5. CQT
7. (X ^ t
1
< t
2
) (.Xxt
1
YYxt
2
) 6. PL
8. (X ^ t
1
< t
2
) (Xxt
1
YYxt
2
) 7. DR4
9. (X ^ t
1
< t
2
) (YYxt
2
.Xxt
1
) 6. PL

181

10. X (Xxt
1
.Xxt
1
) 8-9. PL
153

11. (X ^ Xxt
1
) .Xxt
1
10. PL
12. .!X!x!t
1
t
2
((X ^ Xxt
1
) .Xxt
1
) 11. UG, NEC.
The proof from A6b* has one formal drawback when compared to the argument from the upcoming proof
from A6a*. The shortcoming is that Xxt, where X is a fundamental property, does not entail that x is
permanently X, only that x is X at t in every possible world. The shortcoming is relatively minor, since,
semantically speaking, we would be justified in introducing the following axiom:
Perm.: .(t) !t. (where t is free in the antecedent)
The argument based on A6a*, however will not require Perm. in order to obtain the required result and
hence affords some theoretical parsimony. Accordingly, I will go with the A6a* option (modified to be
restricted to strongly intrinsic properties), although the decision here is relatively arbitrary unless one
believes the case for one of these principles is substantially stronger than the other.
2.3 Lewis Three Responses: Relationalism, Presentism, and Temporal Parts
Lewis (1986a) presents three responses that one might make to PTI, and similar response can be made
with respect to the fixity of fundamental properties, since the latter is simply a more restricted case of the
former.
The first response is to relativize intrinsic properties to times. Indeed, when all atomic formulas
involve times this will be universally the case, except those atomic formulas involving fundamental
properties would hold at all times and worlds. In such cases said property is not relative in any
meaningful sense of the term, but when properties are restricted to some times and worlds but not others,
then they may well be construed as extrinsic. In such cases, however, x being X should not be
understood as extrinsic relative to a time t so much to certain extrinsic relations that x bears with other
objects at t but not at all times.
The second response is to adopt presentism. If only the present exists, it is not the case that both
Ruth is sitting and Ruth is standing are true and hence the problem fades away. Hence the first
premise of PTI is denied. In the modal case, one would deny that there are other possible worlds at all. I
have already, albeit informally, argued that presentism cannot be held in conjunction with HCA.
The third and final option that Lewis discusses, being the option that he favors, is to embrace the
doctrine of temporal parts in the case of temporal change and counterparts in the case of transworld

153
The t
1
< t
2
is dropped out of the antecedent, since t
2
is no longer part of any atomic formulas.

182

variations.
154
By adopting temporal parts, one can say that one part of Ruth is standing at t and a different
part of her is sitting at t. Hence the temporal parts solution denies that strict identity obtain between
distinct times while the counterpart solution denies transworld identity. Whatever other problems one may
have with the metaphysics of temporal parts or counterparts, the former is rendered impossible by the
conjunction of PE and PI while the latter is rendered impossible by the conjunction of NE and NI.
Lewis takes issue with relationalism because he thinks that it would be incredible should it turn
out that all changeable or contingent properties were extrinsic. Indeed, he seems to suppose that this
would render all such properties unnatural. This, again, is based on the Humean assumption that all
natural properties are intrinsic, which assumption I have rejected, since it would seem to require that all
the properties discussed in physics are unnatural. Moreover, I am also making the weaker claim that
fundamental properties are fixed, so intrinsic properties of the compositional variety, provided that there
are any, may yet be temporary and contingent. Moreover, I am not denying that there are intrinsic natural
properties. I am only denying that we are capable of knowing them, a point that Lewis (2009) concedes,
albeit for different reasons.
4. A6 and its Consequences of A6
A6:SX:
S
X !x!t(Xxt (Xxt E!xt)))
Before demonstrating the fixity of strongly intrinsic properties on the basis of A6, the following auxiliary
theorem is required:
T89. !X!x!t((Xxt E!xt) .!tXxt)
1. (Xxt E!xt) .!t(E!xt Xxt) Def
2. (Xxt E!xt) !t(.E!xt .Xxt) 2. K
3. (Xxt E!xt) (!t.E!xt !t.Xxt) 3. CQT
4. ((Xxt E!xt) ^ !t.E!xt)) !t.Xxt 4. PL
5. ((Xxt E!xt) ^ .!tE!xt)) .!tXxt 5. BF
6. .!tE!xt PE, NE
7. (Xxt E!xt) .!tXxt 7. PL
8. !X!x!t((Xxt E!xt) .!tXxt) 8. CQT
Fixity (i.e., the conjunction of necessity and permanence) is established in independently of Perm. in
virtue of the fact that existence has been shown to be fixed independently of Perm..

154
Of course, we have neither change nor variation as one would have originally imagined. No objects really
changes, rather they are replaced by objects which are typically quite similar to their predecessors. Change indeed
happens, but not within objects.

183

T90. Fixity of Strongly Intrinsic Properties: !x!X!t((
S
X ^ Xxt) !t.Xxt)
1.
S
X !x!t(Xxt (Xxt E!xt)) A6, PL
2.
S
X !x!t(Xxt (Xxt E!xt)) 1. PL
3.
S
X (Xxt (Xxt E!xt)) 2. CQT
4. (
S
X ^ Xxt) (Xxt E!xt)) 3. PL
5. (Xxt E!xt) .!tXxt T89, CQT
6. (
S
X ^ Xxt) .!tXxt 4-5. PL
7. !x!X!t((
S
X ^ Xxt) .!tXxt) 7. PL
Whenever an object possesses a strongly intrinsic property X it necessarily always possesses X.
T91: Fixity of Fundamental Properties (FFP): .!X!x!t((X ^ Xxt) .!tXxt)
1. (
S
X ^ Xxt) .!tXxt T90
2. X (X ^
S
X) Def. 60
3. ((X ^
S
X) ^ Xxt) .!tXxt 1. PL
4. (X ^ Xxt) .!tXxt 2-3. PL
5. .!X!x!t((X ^ Xxt) .!tXxt) 4. UG, NEC
T92: Fixity of Fundamental Property Non-Instantiation: .!X!x(tXxt (X
.Xxt))
1. (X ^ Xxt) .!tXxt T91
2. .!tXxt (X ^ Xxt) 1. PL
3. tXxt (X ^ Xxt) 2. Def. ,
4. tXxt .(X ^ Xxt) DR4
5. tXxt .(X \ Xxt) 4. PL
6. tXxt .(X Xxt) 5. PL
7. .(X Xxt) (.X .Xxt) K
8. X .X T18, PL
9. .(X Xxt) (X .Xxt) 7-8. PL
10. tXxt (X .Xxt) 6-7, 10. PL
11. t(Xxt ^ X) .Xxt 10. PL
12. t(Xxt ^ X) !t.Xxt 11. CQT
13. .!X!x(t(Xxt ^ X) !t.Xxt) 12. UG, NEC

184

Necessarily, for every property X, object x, if there exists some time t such that x fails to be X and X is a
fundamental property, then at every time in every possible world x fails to be X.
T93: Fixity of Fundamental Objecthood
!x!t(xt .xt)
1. (X ^ Xxt) !t.Xxt T91
2. (X ^ Xxt) (X ^ !t.Xxt) 1. PL
3. X(X ^ Xxt) X(X ^ !t.Xxt) 2. CQT
4. X(X ^ Xxt) !tX(X ^ .Xxt) 3. CQT
5. X(X ^ Xxt) .!tX(X ^ Xxt) 4. T20, K
6. xt X(X ^ Xxt) Def. x
7. X(X ^ Xxt) xt 6. PL
8. .!tX(X ^ Xxt) .!txt 7. CQT, K
9. X(X ^ Xxt) .!txt 5, 8. PL
10. xt X(X ^ Xxt) 6. PL
11. !x!t(xt .!txt) 9-10. PL, UG
First Corollary: If is the cardinality of the fundamental objects, then the cardinality of the fundamental
objects is necessarily .
Second Corollary: If is the cardinality of concrete objects, then the cardinality of concrete objects is
necessarily equal to or greater than .
For any object x and time t, if x is a fundamental object at t then x is necessarily always a fundamental
object.
T94: Fixity of Fundamental Indiscernibility: .!xy!X!t(((X ^ (Xxt Xyt)
!t.(Xxt Xyt))
1. ((X ^ Xxt) .!tXxt) T91
2. (tXxt ^ X) .Xxt) T92
3. (X ^ (Xxt ^ Xyt)) (.!tXxt ^ .!tXyt) 1, 3. PL
4. (X ^ (Xxt ^ Xyt)) .(!tXxt ^ !tXyt) 3. K
6. ((X ^ (Xxt Xyt) (X ^ ((Xxt ^ Xyt) \ (Xxt ^ Xyt))) PL
7. (X ^ (Xxt ^ Xyt)) (.!tXxt ^ .!tXyt) 4. PL/
8. (X ^ (Xxt ^ Xyt)) .(!tXxt ^ !tXyt) 7. K

185

9. (X ^ ((Xxt ^ Xyt) \ (Xxt ^ Xyt))) (.(!tXxt ^ !tXyt) \ .(!tXxt ^
!tXyt)) 4, 8. PL
10. (X ^ ((Xxt ^ Xyt) \ (Xxt ^ Xyt))) .((!tXxt ^ !tXyt) \ (!tXxt ^
!tXyt)) 9. K
11. (X ^ ((Xxt ^ Xyt) \ (Xxt ^ Xyt))) !t.((Xxt ^ Xyt) \ (Xxt ^ Xyt))
10. CQT
12. (X ^ ((Xxt ^ Xyt) \ (Xxt ^ Xyt))) !t.( Xxt Xyt) 11. PL
13. ((X ^ (Xxt Xyt) !t.( Xxt Xyt) 6, 12. PL
14. .!xy!X!t(((X ^ (Xxt Xyt) !t.(Xxt Xyt)) 13, UG, NEC
Any two objects that are, at any time, indiscernible with respect to their fundamental properties are
necessarily always indiscernible with respect to their fundamental properties.
T95: Necessity of Simple Monadic Fundamental Events: !e!X!t!x(((e = Xxt) ^ (X ^
e)) .e)
1. (e = Xxt) .(e Xxt) T27
2. (e = Xxt) (.e .Xxt) 1. K
3. (X ^ Xxt) .!tXxt T01
4. (X ^ Xxt) !t.Xxt 3. CBF
5. (X ^ Xxt) .Xxt 4. CQT
6. X (Xxt .Xxt) 5. PL
7. .Xxt Xxt T
8. X (Xxt .Xxt) 6-7. PL
9. ((e = Xxt) ^ X) ((.e .Xxt) ^ (Xxt .Xxt)) 2, 8. PL
10. ((e = Xxt) ^ X) (Xxt .e) 9. PL
11. ((e = Xxt) ^ X) (e .e) 10. I2
12. ((e = Xxt) ^ (X ^ e)) .e 11. PL
12. !e!X!t!x(((e = Xxt) ^ (X ^ e)) .e) 12. UG
Accordingly, the monadic fundamental base is fixed across possible worlds, which is simply a result of
the fact that fundamental properties instances are factually equivalent to the existence of their bearers and
existence is fixed.
Next, in order to show that the same pluralities of monadic fundamental events occur in every
possible world, it is first necessary to demonstrate the following theorem of event logic:
T96: !ee(!e((e ee) .e) .ee)

186

1. ee !e((e ee) e) ee5
2. !e((e ee) e) ee 1. PL
3. .!e((e ee) e) .ee 2. K
4. !e(.(e ee) .e) .ee 3. CBF, K
5. !e(.(e ee) ^ .e) .ee 5. PL
6. (e ee) .(e ee) ee4
8. !ee(!e((e ee) .e) .ee) 7. UG
T97: Necessity of Pluralities of Simple Monadic Fundamental Events:
!ee(!e((e ee) (X((e = Xxt) ^ (X ^ ee)) .ee)
1. ((e = Xxt) ^ (X ^ e)) .e T95
2. X((e = Xxt) ^ (X ^ e)) .e 1. CQT
3. !e((e ee) ^ .e) .ee T96
4. ((e ee) (X((e = Xxt) ^ (X ^ ee)) X((e = Xxt) ^ (X ^ e)) PL T32
5. ((e ee) (X((e = Xxt) ^ (X ^ ee)) .e 2, 4. PL
6. ((e ee) (X((e = Xxt) ^ (X ^ ee)) ((e ee) ^ .e) PL
7. !e((e ee) (X((e = Xxt) ^ (X ^ ee)) !e((e ee) ^ .e) 6. CQT
8. !e((e ee) (X((e = Xxt) ^ (X ^ ee)) .ee 7, 3. PL
9. !ee(!e((e ee) (X((e = Xxt) ^ (X ^ ee)) .ee) 8. CQT
T98: Contingent Fundamental Events Involve Fundamental Relations:
!ee((ee ^ (ee ^ .ee)) eX((e ee) ^ X
n
x
1
x
n
t(X
n
^ e = Xx
1
x
n
t)))
1. ee (e ee (X(X ^ (e = Xxt)) \ X
n
x
1
x
n
t(X
n
^ e = Xx
1
x
n
t)))
Def. ee
2. ((e ee) ^ X((e = Xxt) ^ X) ^ ee) .ee T97
3. ((e ee) ^ X((e = Xxt) ^ X)) (ee .ee) 2. PL
4. (ee .ee) ((e ee) ^ X((e = Xxt) ^ X)) 3. PL
5. (ee ^ .ee) ((e ee) ^ X((e = Xxt) ^ X)) 4. PL
6. (ee ^ .ee) ((e ee) \ X(X ^ (e = Xxt)) 5. PL
7. (ee ^ .ee) ((e ee) X(X ^ (e = Xxt)) 6. PL
8. (ee ^ (ee ^ .ee)) ((e ee) X(X ^ (e = Xxt)) 1, 7. PL
9. (ee ^ (ee ^ .ee)) ((e ee) X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t))) 1, 8. PL
10. (ee ^ (ee ^ .ee)) !e((e ee) X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t))) 9. CQT
11. (ee ^ (ee ^ .ee)) e((e ee) ^ X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t)))

187

10. ee3, CQT
12. !ee((ee ^ (ee ^ .ee)) e((e ee) ^ X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t))))
11. UG
T99: Contingent Truths Require Fundamental Relations:
.( ^ .) ee((ee ^ ( ee)) ^ e((e ee) ^ X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t))))
1. ee(ee ^ ( ee)) A3
2. ee(ee ^ .(ee )) 1. Def.
3. (.(ee ) ^ .) .ee PL
4. ( ^ .) ee((ee ^ ( ee)) ^ .ee)) 2-3. PL
5. (ee ^ (ee ^ .ee)) e((e ee) ^ X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t)))T98
6. ee((ee ^ ( ee)) .ee)) ee(ee ^ (ee ^.ee)) F10
7. ee(ee ^ (ee ^ .ee)) e((e ee) ^ X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t)))
5. CQT
8. ( ^ .) e((e ee) ^ X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t))) 4, 6-7. PL
9. ( ^ .) ee((ee ^ ( ee)) ^ e((e ee) ^ X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t))))
1, 8. PL
If is contingently true, there is some fundamental plurality of events such that is factually equivalent
to ee and there is some e contained in ee that is identical to the instantiation of some fundamental (and
hence natural) relation. This is a particularly important theorem in that it links all contingent property
instantiations to the natural relations that obtain between fundamental objects.
T100: Contingency Entails Object Plurality: ( ^ .) .x
1
...x
n
((x
1
^x
n
t)

^
x
1

t
x
n
)
1. (ee ^ e((e ee) ^ X
n
x
1
x
n
t (X
n
^ (e = Xx
1
x
n
t)))) ex
1
...x
n
((x
1
^x
n
t)

^ x
1

t
x
n
) Def. X
n
, X
n

2. ee((ee ^ e((e ee) ^ X
n
x
1
x
n
t (X
n
^ (e = Xx
1
x
n
t))))) ex
1
...x
n
((x
1
^x
n
t)
^ x
1

t
x
n
) 1. CQT
3. ( ^ .) ee((ee ^ ( ee)) ^ e((e ee) ^ X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t))))
T99
4. ( ^ .) ee((ee ^ e((e ee) ^ X
n
x
1
x
n
t(X
n
^ (e = Xx
1
x
n
t))))))
3. PL
5. ( ^ .) x
1
...x
n
((x
1
^x
n
t)

^ x
1

t
x
n
) 2, 4. PL
6. x
1
...x
n
((x
1
^x
n
t)

^ n > 1) .x
1
...x
n
((x
1
^x
n
t)

^ x
1

t
x
n
) T97, T2

188

7. ( ^ .) .x
1
...x
n
((x
1
^x
n
t)

^ x
1

t
x
n
) 5-6. PL
If any formulas are contingently true, then there must be a plurality of fundamental objects.
T101: ( ^ .) .!t!xy((xt

^

yt)

x =
t
y)
1. ( ^ .) .x
1
...x
n
((x
1
^x
n
t)

^ x
1

t
x
n
) T100
2. .x
1
x
n
((x
1
t^x
n
t)

^ x
1

t
x
n
) .xy((xt

^

yt)

^ x
t
y) CQT
3. .xy((xt

^

yt)

^ x
t
y) .xy((xt

^

yt)

x =
t
y) 2. PL
4. ( ^ .) .xy((xt

^

yt)

x =
t
y) 1, 3. PL
5. ( ^ .) .!xy((xt

^

yt)

x =
t
y) 5. Def
6. ( ^ .) .!t!xy((xt

^

yt)

x =
t
y) 5. UG, BF
T101 effectively says that if there are any contingent truths, then priority monism is false.
T102: Denial of Priority Monism: .!t!xy((xt

^

yt)

x =
t
y)
1. (ee ^ .ee) .!t!xy((xt

^

yt)

x =
t
y) T101 ee/
2. !ee(ee .ee) A1
3. ( ^ .) !xy((xt

^

yt)

x =
t
y) 1. T, CQT
4. (ee ^ .ee) !xy((xt

^

yt)

x =
t
y) 1, 3. PL
5. !xy((xt

^

yt)

x =
t
y) (ee ^ .ee) 4. PL
6. !xy((xt

^

yt)

x =
t
y) !ee(ee ^ .ee) 5. CQT
7. !xy((xt

^

yt)

x =
t
y) !ee(ee ^ .ee) 6. PL
8. !xy((xt

^

yt)

x =
t
y) !ee(ee .ee) 7. PL
9. !ee(ee .ee) !ee(ee .ee) DR5
10. !xy((xt

^

yt)

x =
t
y) !ee(ee .ee) 8-9. PL
11. !ee(ee .ee) !xy((xt

^

yt)

x =
t
y) 10. PL
12. !xy((xt

^

yt)

x =
t
y) 2, 11. PL
13. .!t!xy((xt

^

yt)

x =
t
y) 12. UG, NEC
Corollary: Denial of Existence Monism: .!t!xy((Cxt

^

Cyt)

x =
t
y)
The denial of existence monism follows upon the denial of priority monism: if there is a proper
plurality of fundamental object then there is proper plurality of concrete objects.
While existence monism is not a widely held view, it does have some exponents, the earliest
explicit one being Parmenides. Its denial is rendered somewhat significant in the presence of the fixity of
existence. If universe as a whole were the only real substance, and necessarily so, then the fixity of
existence would follow as a matter of course. Every world would certainly have the same domain of

189

objects, namely the universe itself. Moreover, it is evident that the universe would endure for the whole of
time (or rather the extent of time would be explained by the endurance of the universe). Indeed, given
monism, one could literally follow Stalnakers (1973, 2011) suggestion that possible worlds should be
construed as maximal properties or ways the world might have been. Accordingly, the fixity of existence
would be no more controveral than existence monism as such and the fact that the fixity of existence is a
consequence of classical logic would seem to offer existence monism some vindication. Moreover, given
monism and HCA, the fixity of existence becomes quite plausible independently of SQML. That is, if all
possibility is grounded upon ways in which the actual world might have become (from any arbitrary time
in the past) it would be strange if the universe contingently evolved in such a way that it turned out to be
the only concrete object (i.e., such that it did not possess any proper parts at all).
The possibility of existence monism, however, is effectively blocked by A6 in conjunction with
A1. In other words, while existence monism is compatible with SQML as such, it would not seem to be
compatible with the conjunction of Strong Nominalism and SQML. That is, monism plus SQML would
seem to require in rebus universals or tropes.
As for priority monism, I have construed it roughly along the lines of Shaffer (2009), who takes
the non-fundamental to be nothing over and above the fundamental, which means, from a metaphysical
point of view, priority monism is just a way of trying to make sense of existence monism. It remains
possible, however, that there is a first cause of everything other than itself, provided that it necessarily
causes all other substances to exist.
Moreover, since I have not identified the fundamental with the physical, T116 is not incompatible
with physical monism: the doctrine that there is only one physical object. It might still be the case that
there is but a single and continuous material substratum and that the diversity exhibited within it is
somehow a reflection of the multiplicity of non-physical fundamental objects and their relations.
T103: Relationality of Contingent Natural Properties
. !X!x!t(((Xxt ^ .!tXxt) ^ X) X)
1. X (X ^ X) Def. X
2. ((X ^ Xxt) .!tXxt) T91
3. (X (Xxt .!tXxt) 2. PL
4. (Xxt .!tXxt) X 3. PL
5. X (X ^ X) 1. PL
6. X (X \ X) 6. PL
7. (Xxt .!tXxt) (X \ X) 4, 6. PL

190

8. (Xxt .!tXxt) (X X) 7. PL
9. (Xxt ^ .!tXxt) (X X) 8. PL
10. ((Xxt ^ .!tXxt)) ^ X) X 9. PL
11. !X!x!t(((Xxt ^ .!tXxt) ^ X) X) 10. UG
This result was already, in principle, established with T113, but is here made explicit in the interest of
good housekeeping.
5: The Necessity of Fundamental Properties and Saving the Phenomena
One might wonder how it could even be the case that all change could be relational. How can
this claim to square off with the phenomena? Let us assume for the moment that natural relations are all
extrinsic in nature. In the case of properties that depend upon global fields, it would appear that each such
field is externally related to every observable object. Such ubiquitous external relations, however, only
ground contingent parameters, such as the actual laws of physics, the fundamental forces, and which
particle kinds can be instantiated. There might also be a variety of quasi-global conditions, some of which
might appear to be global because they envelop the entire observable universe, but not the entire existent
universe. Or perhaps some galaxies or galaxy clusters are subject to conditions specific to those regions,
such as galaxy-wide magnetic fields. More locally, our ordinary or pre-theoretical conceptions of objects
presuppose conditions which are quite literally global, namely those which are particular to conditions
which are ubiquitous on the planet Earth or at least in those regions that are inhabited by humans. Such
conditions involve a specific g-force, electromagnetic field, atmospheric pressure, along with the overall
chemical makeup of the lower atmosphere. Accordingly, when we consider ordinary objects there are a
wide variety of background conditions that we hold fixed. These conditions are sufficient for our
judgment that all observable properties are extrinsic, but they alone do not explain why the observable
properties of objects change over time in accordance with observation.
Let us try to make this difficulty more precise. It should be noted that we need not necessarily
bother ourselves with changes in composite objects here, since their changeable properties can all be
understood to be compositional, (unless they are emergent but I shall put off discussing this possibility
for the moment). Hence, the only phenomena we need to save belong not to the manifest image but to the
scientific image. This concern was partly addressed when it was observed that the physical properties of
simple particles all appear to be extrinsic, and hence non-fundamental. But even if no observable
properties are fundamental, it might still be the case that some changes in non-fundamental properties
may be in virtue of changes in fundamental properties. Observable properties are, after all, only partly
grounded in natural relations, they are also partly grounded in fundamental properties. Changes in non-

191

observed fundamental properties would, in conjunction with the various extrinsic factors, cause changes
to occur in the extrinsic properties that we observe. Moreover, this seems to be highly plausible due to
the fact that extrinsic factors must remain constant in order to have their effect. The Higgs field permits
particles to have mass, but interaction does not permit the continued possession of mass after the
interaction has ceased to take place. The interaction must be preserved so long as the particle in question
is to have mass. It is often the case, or so it seems, that a state resulting from a local interaction persists
long after said interaction. So it would seem that we require fundamental change after all.
This objection, however, assumes that external relations are always, or typically, local. Perhaps
brief local external relations frequently result in lasting non-local relations that preserve changes in
extrinsic natural properties. Indeed, entanglement relations clearly satisfy all of these conditions except
for the last and they could be understood to satisfy the last as well. An entanglement relation does not
bestow any determinate observable properties upon its relata, but it does alter the dispositions of its relata
and hence it must affect the categorical state of its relata, even if such properties happen to be elusive. So
it is not too farfetched to suppose that such relations do sustain extrinsic properties possessed by their
bearers.
Entanglement relations alone might not sufficiently explain the appearance of persisting change
resulting from relatively punctual and local interactions, but, if not, the sort of relations needed will at
least be similar to entanglement relations and hence are not as outlandish as one might have thought. On
this hypothesis, we might picture local interactions as if they resulted in adding and removing lines
linking distinct objects. When objects x and y are locally and briefly X
2
related they might come to be Y
2

related for a period of time corresponding to the time during which x or y posses some extrinsic natural
property. The event of the coupling may also de-couple x or y from some objects with which x or y had
earlier been coupled. This would be in virtue of certain incompatibilities between various sorts of
couplings. For any object x, the sum total of these lines or couplings at any given time would determine
those extrinsic natural properties it possesses which are not had in virtue of general environmental factors.
One might object to a metaphysical theory that may need to posit a new class of natural relations
which are, apparently, not posited by experimental science. It might be said, however, that we are already
doing this when posting intrinsic natural monadic relations (i.e. properties), since natural science does not
seem to require such properties. Objections to the existence of intrinsic properties are relatively
uncommon. What makes it worse to posit natural relations? Sure, it violates Humean intuitions regarding
the complete independence of distinct entities, but this entire treatise has been anti-Humean in spirit. In
positing intrinsic natural properties we are supposing that there is some deeper level of qualitative reality
about which physics is silent; in positing new natural relations we are supposing that there is some deeper

192

level of connectivity between objects about which physics is silent. Indeed, there is more hope that
physics will assist us in discerning the fundamental relations than in discerning the fundamental
properties. This is because there is truth in the claim that science only really teaches us about the
structures of things, which structures can be mathematically represented. Pure qualities, qualities that are
not factually equivalent to compositional properties, are mathematically elusive and irredeemably so. This
is precisely why quality cannot be adequately defined by model theory it must be taken as completely
primitive. Relations, however, lend themselves to mathematical representations and so it is epistemically
possible that the sort of relations we need may be posited in future physics. Moreover, since there are
current physics relations that are similar to those which we need (and may even be sufficient), the
prospects here are not entirely dim.
Finally, one might recall our brief encounter with the position of radical structuralism in Chapter
One. Said structuralism, also known as ontic structural realism (OSR) denies that objects possess any
intrinsic properties at all. Hence, if OSR should be phenomenological adequate, it would follow that
changes in extrinsic properties will be adequate in explaining physical change. All that I am adding is the
claim that intrinsic properties underlie the very capacity to enter into such relations in the first place.
Accordingly, if OSR can save the phenomena with respect to physics my own position can do so just as
easily.
More would need to be said regarding how it is that we might save the phenomena on the basis of
relational change alone. I only wanted to do two things here:
1. Make the problem clear by indicating that the relation (compositional or extrinsic)
underlying an instance of a extrinsic natural property X will have to remain constant (or near
enough) so long as said instance of X is retained.
2. Attempt to persuade the reader that this problem is not hopeless and hence does not render
the theory a whole (or rather the theory including A6) a non-starter.
This problem, however, cannot be fully engaged until other details of the theory have been fleshed out. In
particular, we have really said nothing decisive about the world of appearance as such. Accordingly, we
cannot yet say how fundamental reality is to be related to either the manifest image or, for that matter, the
scientific image.
Of course, it is always possible that the fixity of fundamental properties will turn out to be simply
untenable relative to any plausible way of connecting a theory of fundamental objects with any direction
physics is likely to turn. Hence, it is surely possible that A6 will eventually have to be dropped for
empirical reasons. My primary reason for positing A6 was because it provides a simple and elegant way

193

of rounding off the general system of necessary substances. The next, and final, chapter will be partly
dedicated to exhibiting some of the resulting symmetries.
But perhaps such rounding off is not such a good thing. Perhaps it is a symptom of artificiality,
as many would deem to be the case with the Generalized Continuum Hypothesis in set theory. In that
case, one might consider theorems based upon A1-A5 to be plausible while refusing to accept those that
require A6. On the other hand, perhaps A6 will turn out more like the Axiom of Choice, initially
counterintuitive but eventually accepted as an obvious truth. The system extended to A6 is, however, in
the very least, intellectually interesting and so it might be worth considering if only to engage in a kind of
intellectual hedonism.
Before proceeding, however, I will note that I can conceive of one alternative way in which the
present axiomatic metaphysics might have been rounded off. That would have been to deny that there are
irreducibly external relations at all, which would mean that there could be no extrinsic properties. This
alternative is clearly incompatible with A6 (given necessitism and A1). Given the substance ontology
inspiring A2-A5, however, one might well wonder just how irreducible external relations are possible in
the first place. This is especially so given the fact that permanentism will not permit substance-like
spacetime points within ones ontology. Nevertheless, it would be challenging to reducing all external
relations to internal relations. We would have to understand change in distance between two objects in
terms of changes in their intrinsic properties. The problem can be elevated slightly if the intrinsic relation
need not involve precisely the objects in question:
Do32: Monadism: !x
1
x
n
!X
n
!t((X
n
^ X
n
x
1
x
n
t) y
1
y
m
Y
m
(Y
m
^ (X
n
x
1
x
n
t
Y
m
y
1
y
m
t)))
Modadism is the doctrine that any external relatons are grounded in some internal relation.More
economically, I might have required that only monadic events may count as fundamental; in which case
A3 would inevitably lead to something along the lines of monadism.
Monadism is the doctrine all putatively external relations are grounded in internal relations. The
grounding internal relation might include all of the objects in the grounded external relation, but it might
include others besides.
It is difficult to conceive how this might work, but Heil (2013) suggests one possibility. Suppose
that the whole of space is a single extended substance, as Spinoza believed.
155
What we call individual
material substances would really be thickenings in certain regions of this one continuous space.

155
Heil primarily attributes this view to Descartes.

194

Objects, in this respect, would resemble illuminations on your computer monitor. The motion of
an object would resemble the motion of lights around a theatre marquee. The deep story about the
motion of the billiard ball across the table would be that, at the fundamental level, nothing moves.
Better, truthmakers for claims about moving objects people, automobiles, footballs would be
non-moving features of the universe. If this sounds odd, think of the truthmaker for a judgment a
true judgment that your cursor moves from one region of the screen to the other (Heil 2012:
147).
Relative to current physics, Heil suggests that one might replace space with the global fields of
field theory. In this case, however, the individual particles will be regarded as excitations in these fields,
and hence reduced to mere modes. Spatial distances between excitation are intrinsic to the fields. Hence
we begin to move toward something similar to monism, especially if all of the global fields are unified
at some fundamental level. Hence, opting for Monadism rather than A6 would have resulted in a theory in
which the prospects for monism would have been quite strong indeed. While I do find Monadism
appealing, I have opted for A6 instead as a means of rounding off the total theory.
















Copyright Adam Labecki 2013

195

CHAPTER TEN:
OBJECTS, PROCESSES, OUTSTANDING QUESTIONS,
AND CONCLUDING REMARKS


1. Introduction: the Beginning of the End
This treatise has presented an axiomatic metaphysics of substances. While the discussion portions have
sometimes drawn upon natural science and some basic tenants of ordinary knowledge, the formal
axiomatic theory only concerns itself only with the conditions that must be satisfied in order for
something to count as a substance. It is quite silent on just what the substances are, much less does it tell
us which, if any, identifiable substances are fundamental. Indeed, it does not tell us whether there are any
non-fundamental substances at all. The logical axioms in conjunction with A1-A6 are perfectly
compatible with an empty class of non-fundamental entities. It could be that there are no composite
substances or non-fundamental emergent substances. More generally, the location problem has been
largely put out of play. We have, however, come to a number of unexpected results arising from logic and
some mostly non-controversial axioms (A6, while motivated, remains fairly controversial). These results
place certain restrictions on how location problems might be addressed and perhaps even on our overall
approach to such questions.
In this final chapter I shall consider how the division between fundamental enduring objects, on
the one hand, and occurent processes, on the other, might assist in connecting the present theory to our
ordinary understanding of the world. The connection, however, will be conditional upon the assumption
that some subset of the objects we normally name (or could name) correspond to genuine substances.
Inasmuch as any ordinary objects correspond to substances at all, they will be processes of a special kind
that I shall call lifetimes (although not all such processes necessarily involve organic beings).
Accordingly, a certain division between objects and processes, Being and Becoming, will be discerned.
When attempting to identify the fundamental objects, we need not confine ourselves to the
objects of fundamental physics. A number of similarities will be discerned between the present theory and
Leibnizian monadology, on the one hand, and certain sects of Hinduism, on the other. One feature
common to both of these but on which the present treatise has been silent, is panpsychism (ire., the
doctrine that all fundamental beings possess some form of mentality or consciousness).
As for the Being and Becoming distinction, its relevance to the consideration of standard location
problems will largely depend upon whether there are any composite substances or emergent properties. I

196

shall not have too much to say about the prospects of such items, expect to say that composite substances
(or something near enough) would presuppose some form of ontological emergence (but it may be the
case that relatively non-exotic forms of emergence would suffice). The treatise ends with a brief
suggestion regarding the nature of organic substance.
2. On the Lifetimes of a Substance
Nelson (2007), when discussing BPAE, suggested that we treat natural sortals as modes of an object,
roughly on par with other modes or accidents. Introducing modes here is a good start, but we will want
to be able to refer not only to such phase sortals as such but to phase sortal instances or tropes. I write
tropes in quotation marks because I do not literally believe that concrete particulars are constituted by
abstract particulars in any way whatsoever (after all, I was just endorsing Strong Nominalism in the last
chapter). That is, I do not believe in in rebus tropes but I am perfectly okay with post rem tropes, which
are abstract objects, or products of analysis. If I, or anyone, can conceive of it, it at least exists as a
conceptual object. Now I do want to get at a conceptual object that nevertheless cuts at some natural
joints. Cutting at the joints, however, does not require that all the cuts are made between things.
Suppose that each human x living at any given time t is identical with some substance y for which
human is a phase sort instantiated by y (and hence also x) at t. If one believes that humans could not
even count as non-fundamental substances on the present theory (perhaps because necessary composite
objects seem unlikely) then pick whichever natural kind sortal you please.
The name Aristotle, in as much as it refers to a necessary and hence permanent substance, does
not refer to a homo sapiens lifetime that begins in 384 BC and ends in 322 BC. Let us call Aristotle the
substance Aristotle
S
and Aristotle the lifetime Aristotle
L
. Aristotle
L
is a process that strongly involves
Aristotle
S
during a span of time that begins in 384 BC and ends in 322 BC. In ordinary language the
meaning of the name Aristotle involves many of the properties that the process Aristotle
L
strongly
involves (i.e., properties that are had by Aristotle
S
throughout the entire span of Aristotle
L
). Logically
speaking, however, Aristotle, inasmuch as it is names a member of D, must be understood to mean
Aristotle
S
, it just so happens that we rely on Aristotle
L
in order to fix reference to Aristotle
L
. If Aristotle
S

happens to be identical with Locke
S
then both Locke
L
and Aristotle
L
are distinct lifetimes of one and the
same substance.
Recall that, by T40, for each object x and time t there exists an event ee that includes each event
that involves both e and t (at the world of evaluation). We will want to have a more fine-grained sort of
event, however, when beginning to characterize lifetimes.

197

Def100: x-Characterizing Event at t: ee is an x-characterizing event of x at t, written ee[x, t],
just in case:
!e((e ee) (((e ^ et) ^ (ex) ^ !X(eX (X ^ Xxt)))
An event ee characterizes x at t just in case it includes each natural property instantiation involving x at t
and no other simple events.
T104: !x!tee!e((e ee) (((e ^ et) ^ (ex) ^ !X(eX (X ^ Xxt)))
As with other similar theorems, this is effectively a result the ee2 comprehension schema. For every
object x and time t there is exactly one x-characterizing event at t.
Def101: x-Characterizing Process: ee
1
is an x-characterizing process during time span t
1
t
n
,
written ee
1
([x]t
1
...t
n
), if and only if:
!ee
2
((ee
2

t
ee
1
) (ee[x, t
1
] \ ee[t
n
]))
An x-characterizing is a process, each stage of which is an x-characterizing event, for some object x.
T105: !x!t
1
t
n
(St
1
t
n


ee
1
!ee
2
((ee
2

t
ee
1
) (ee[x, t
1
] \ ee[x, t
n
])))
This theorem follow on the basis of Pee1 in conjunction with T27, which guarantees that, for each x and
each span t
1
t
n
, there is always some thin process along t
1
t
n
strongly involving x.
If a is the object underlying Aristotle
L
(i.e., a = Aristotle
S
) then Aristotle
L
will be the natural thick
process of a-characterizing events between (roughly) 384 and 322 BC. Intuitively, a lifetime of an object
x is a natural process of x-characterizing events during which x instantiates the same natural kind sortal.
This characterization, however, will not be sufficient. Suppose that, as soon as Aristotle
L
died (came to an
end) in 322 BC Aristotle
S
is immediately born to another human life, whose surname is Wong. In this
case Wang
L
and Aristotle
L
will be the same with respect to both their underlying substance and the
involved natural kind, since each is a human life. Moreover, there is no temporal gap between Wang
L
and
Aristotle
L.
But it seems that they should count as distinct lifetimes and, inasmuch as there are such things
as persons (in a roughly psychological sense of the term), Wang
L
should count as a different person than
Aristotle
L
(assuming that Wang
L
does not typically remember being Aristotle
L
). The question is: in virtue
of what difference shall Wang
L
and Aristotle
L
be distinguished as two distinct human lifetimes rather than
two halves of a single human lifetime?
Perhaps there is some spatial continuity between each moment of Wong
L
and each moment of
Aristotle
L
but between the first moment of the former the last moment of the latter there is a sudden
spatial jump to a relatively distant location. The spatial continuity thesis, however, would preclude the

198

possibility of a physical substance being teleported to a distinct location while continuing the same
lifetime. More generally, it would not permit objects to make relatively large jumps from one location
to another. This, however, would appear to be a nomological possibility (and especially a wide-
nomological possibility) and so the spatial continuity solution is too restrictive.
I shall analyze the distinction between one lifetime of an object and the next in terms of a
difference in sortal tropes, where the analysis of tropes will itself be reductive. It is natural to represent
ordinary tropes in terms of simple states of affairs or Kim-like events, but, given endurentism, we do wish
to distinguish the two. For instance, for any property kind X, object x, and distinct times t, t, Xxt and
Xxt will count as two distinct Kim-style events but they need not represent two distinct instances of
X. That is, between two consultative moments, an electron may retain the same charge not only in that it
has the same kind of charge at each instant but that the same charge instance is possessed between the two
moments of time. Trans-temporal trope identity may be represented as a quasi-factual equivalence
relation
X
, such that Xxt
X
Xxt reads xs being X at t is X-factually equivalent with as xs being X
at t.
156
Continuity of being X is not sufficient for it being the same instance of X, since whatever was
making x X at t might cease to be the case by t, but immediately replaced by something else making x X.
In the case of fundamental properties, continuity should ensure sameness of instance, since they are
always had in virtue of the existence of their bearer. In the case of relational properties, compositional or
extrinsic, the natural relation(s) that makes it the case that x is X at t will have been retained at t.
Accordingly, if Xxt
X
Xxt and if Xxt and Xxt , then from one obtains simply by
replacing every instance of t with t and from one obtains simply by replacing each instance of t with
t:
Def102: Trope Equivalence
(Xxt
1

X
Xxt
2
) ((( Xxt
1
) ^ (Xxt
2
)) (((t
2
/t
1
) ) ^ ((t
1
/t
2
) ))) ^ !t
3
(t
1
< t
3
< t
2

( Xxt
3
^ ((t
3
/t
1
) )))).
That is, two property instances at time t
1
and t
2
are trop-equivalent just in case they factually
differ, at most, in the time at which they occur and, for every t
2
between t
1
and t
2
, the same property is
also possessed and factually differs from the possession at t
1
, at most, in the time at which they occur.
So by constancy of a trope X of x all I mean is constancy of the time-independent reasons (i.e.,
states of affairs) for x being X, I do not mean to say that there is some item, a trope, that assists in

156
The advocate of PTI-style arguments might wonder if premise two of PTI might be replaced by E!xt
E!

E!xt, interpreted to indicate that x exists at both times for the same reasons. This, however, will be of no good for
PTI unless it is already presupposed that the existence of x depends upon the possession of all of its intrinsic
properties.

199

constituting x at two different times. Given A6, such constancy will be constancy of certain natural
relations while X is x. Since ersatz tropes will be partially explained by constancy of natural relations,
we may say that they do cut at natural joints despite the fact that mind independent tropes, as such, do not
exist.
Df103: Lifetime: Where ee
1
is a process characterizing x during t
1
t
n
, i.e.,
ee
1
([x]t
1
...t
n
), and K a natural kind term, ee
1
is a K-lifetime of x, written ee
1KxL
, just in case
there is a thin process ee
2
embedded in ee
1
satisfying the following conditions:
1. ee
2
[K]
2. !e
1
!e
2
(((e
1
ee
2
) ^ (e
2
ee
2
)) (e
1

K
e
2
))
3. !e
1
((e
1
ee
2
) !e
2
((e
1

K
e
2
) e
2
ee
2
))
That is to say, a K-lifetime of x is an x-characterizing process ee
1
within which there is an embedded thin
process ee
2
that is maximal for xs possession of a given K-trope, for some sortal property K.
While I have renounced any commitment to genuine tropes, I have had to make recourse to the
notion of a natural sortal or natural kind. This notion came up earlier when discussing sortal essentialism,
but it has not been defined in terms of our primitives. Is it necessary that we should arrive at a rigorous
definition of natural kind? That depends upon whether or not we wish to be realists about natural kinds.
If the purpose of this analysis of lifetimes is simply to indicate one way in which ordinary language might
be connected with the metaphysics of necessary substances, then we might well accept the primitiveness
of natural kind, since it will be a term of convention rather than a term of metaphysics per se and we
have, as far as I can tell, no more work to do on the matter.
Conventionalism regarding natural kinds is relatively attractive if only because it is very difficult
to arrive at a definition of natural kind. In the literature there are three major approached to natural kind
realism, each corresponding to a role that natural kinds might play in our theorizing: (1) induction, (2)
laws of nature, and (3) causal explanation.
157
In their immediate form, I find all of these methods to be too
close to conventionalism, precisely because they relegate natural kinds to roles they play in our theorizing
rather than directly characterizing the objective nature of natural kinds as such. This is also the case for
Wiggins (1980/2001) analysis of sortals terms in general, since it is concerned only with counting and
tracking objects. It may have been noticed that the object language developed in this treatise is very much
an object language in that it does not permit us to articulate propositions about theories as such as such.

157
For a review article on the question of natural kinds see Koslicki (2008).

200

Nevertheless, the way of induction naturally lends itself to defining natural kinds as something
like sets of natural properties. This is probably the most obvious and straightforward reductivist approach
to natural kinds. On this view, for each natural kind K, object x and time t, x is K at t will be true just in
case x possesses certain natural properties. Hence, for each natural kind K the essence of K would be a set
of natural properties. Hence, once one knows that x is a K, if G is a member of the essence of K, then one
may infer that x is G.
But alas, this approach to natural kinds is problematic. In biological species, at least, individuals
alter significantly between birth and death. There are also differences between members of a species,
and any given species changes over time. Moreover, members of a species may find themselves deformed
by injury or disease while remaining members of said species. Hence, any set of properties liberal enough
to include every member of a species it liable to be liberal enough to include members of distinct species.
Reductive accounts of natural kinds in this case will also render the distinction between being and
becoming of little use for addressing location problems. If lifetimes, so defined, do not pick out anything
that cuts nature at the joints, then why believe that each such lifetime is the manifestation of a particular
substance?
Are there any non-reductive accounts of natural kinds that might fare better? What about non-
reductive Aristotelian approaches? For an Aristotelian, it is not so much that a natural kind has an essence
as it is an essence. That is, if K is a natural kind and x is K at t, we might say that there are bound to be
natural properties which are had by x at t partly in virtue of the fact that x is K at t. Of course, given
necessitism, we cannot say that a temporally instantiated natural kind is ever the essence of an object as
such, but perhaps a relative or auxiliary essence of an object during a span of time, and it is just such an
essence that characterizes a lifetime.
In the case of biological kinds natural kind would naturally be associated with possessing cells
propagating a certain genetic structure providing an underlying causal mechanism. That is, given the
genetic structure of x at t, some of the natural properties possessed by x will be possessed partly in virtue
of said genetic structure. Indeed, one might suggest that it would be a partial cause for all contingent
natural properties that x possess at any time involved in ee
KxL
, where nourishment, environment, and
other external relations would make up the remaining contingent factors. Even when a creature has a stub
instead of a full leg, the having of the stub is partly in virtue of the way the creature grew prior to loosing
part of its leg.
158


158
Samuels (2012), in the context of discussing human nature endorses a causal essentialist conception of
natural kinds along these lines.

201

While Aristotelianism so described avoids many of the problems of reductive essentialism while
offering a more explanatory theory, it must still address genetic variation of a species over time. The
problem is that we are bound to run up against vagueness where one species ends and the next begins.
159

It seems that the problem of vagueness from evolution is likely to plague any analysis of natural kinds,
reductive or otherwise. Those who think that arguments from vagueness are decisive in other areas of
metaphysics, such as material composition, will probably find it to be decisive here too. Those who are
unimpressed with arguments from vagueness elsewhere will be unimpressed here as well.
Finally, some claim that species should not be analyzed in terms of kinds with members but
rather as individual objects (or populations), with parts rather than members. The population approach
tends to consider not only similarity of DNA but also of interbreeding between members of a population
and homogeneous features of the environmental pressures they face. Some such theories also seek to
positively explain the genetic variation within each species (i.e., one is concerned not only with traits
typically found throughout a species but with the particular variations found among particular groups or
individual members). Despite these advantages, however, the above mentioned problem of vagueness
does not seem to go away. The notion of lineage is introduced, which assists in the individuation of
species, but it does not negate it the problem of vagueness, since we are all descendants of a great many
species, presumably going back to prokaryotes. When tracing back from one homo sapiens ancestor to
another, one would (if one could actually track such a lineage) eventually come to beings who are no
longer homo sapiens, but which ancestor counts as the first non-homo sapiens? Given the slow nature of
genetic mutation and natural adoption, it seems that we face a typical sorties paradox here.
So where does this leave us? We might avoid vagueness by only accepting the lifetimes of
particles, atoms, and molecules. If we wish to include biological entities, however, we need way to
respond to the problem of vagueness. I shall leave these matters open for the present, since they are
controveral but I tentatively suggest that the realist accept some form of Aristotelian realism about natural
kinds, which might be expressed as:
Df104. X is a natural kind property, written X, just in case:
!x!t
1
(Xxt
1
!Y(((Y ^ Yxt
1
) ^ t2Yxt
2
) D
(ccm)
((D
(ccm)
xt
1
Xxt
1
) ^ (Yxt
1
m))))


A natural kind property X is such that, when instantiated, any other property that is temporally
instantiated at that time is a manifestation of some dispositional predicate possessed in virtue of the

159
It is doubtful that this vagueness would be sufficiently undermined by ordinary criteria, such as the ability to
produce sexually fertile offspring. That is, as two species begin to diverge, one would expect a gradual decline in the
capacity for successful reproduction of fertile offspring between members of the two diverging groups. It would be
peculiar if members of the two groups could interbreed perfectly well up until a certain generation, at which point
they were unable to produce fertile offspring with one another.

202

possession of X. It should be noted that this definition is compatible with objects having more than one
natural kind at once. For instance, it could be that, for any object x there is some underlying natural kind
that x instantiates for the entire duration of its existence, and hence the entire duration of time. We do not,
of course, know what those kinds would be or whether there would but one such kind or many. This does
not present a difficulty for the above definition of lifetime, however, since the natural kind in question is
stipulated as being instantiated during a specified span of time. In cases in which the natural kind is
necessarily and permanently instantiated, we are really speaking of the essence of an individual.
It should also be noted that this definition does not address questions regarding how distinct
natural kinds are individuated. Accordingly, I am presenting a definition of natural kind but by no means
a theory or analysis of natural kinds. I will only suggest that, given the genetic nature of natural
organisms it might be better to analyze natural kinds in terms of dispositional ascriptions rather than sets
of essential properties. That is, each natural kind would be understood a causal mechanism underlying a
set of characteristic dispositional ascriptions.
With respect to lifetimes themselves, both sortal essentialism and origins essentialism seem to be
appropriate, since we would not be speaking of the same life if it were of a different natural kind or if it
had different origins. By ee4 all plural events whatsoever (including all processes) necessarily include just
those simple events that they actually include and by T33 all events necessarily involve just those
properties, objects, and times they involve. Accordingly, the only contingent thing about any processes,
and hence about any lifetime, is whether or not it occurs.
Since lifetimes are already very much like four-dimensional objects, we might reasonably
suppose that their modal properties are best understood in terms of counterparts rather than transworld
identity (four-dimensionalists tend to endorse counterpart theory). The object language presented in this
treatise is not adopted to accommodate such relations, but might be modified to do so if it proved useful.
For any given K-lifetime, ee
KxL
, of an object x, the counterparts of ee
KxL
would be precisely those K-
lifetimes of x that share in the same origin. In this respect the modality of lifetimes would be more similar
to that of worlds than objects, at least given the branching-history account of modality that I have
provided in Chapter Eight. Each contingent ee
KxL
would be like a branch on sub-tree of x-possibilities.
Each of these counterparts concern events which are already metaphysically possible for x but we have
restricted ourselves to those x-possibilities that that branch off from the moment at which x becomes a K
at a certain time t, each one persisting in its respective world, w, until x ceases to be that very K in w.
160


160
One might, of course, admit some degree of flexibility to the matter of origins. In particular, one might admit
flexibility regarding the timing of the initial origin stage.

203

As such, each K-lifetime of an object will strongly involve K in every possible world, have the same
origin in every possible world, and is liable to have additional non-trivial essential properties besides.
As mentioned above, lifetimes are similar to four-dimensional objects. Both are, roughly
speaking, occurents (processes) composed of smaller occurents (events). That is, each can be said to be
composed of stages as minimal temporal parts, as well as segments, being sets of stages without any
temporal gaps.
There are, however, important difference between lifetimes and the spacetime worms of four-
dimensionalism. It was already noted that, on the present view, a trope theorists could retain trope
identity between distinct times, but this would not be so on four-dimensionalism, where there is never any
strict identity between anything between two distinct times. Moreover, four-dimensionalism encourages a
fundamental ontology of spacetime points, while the present system forbids this. Where four-
dimensionalism effectively replaces endurents with purdurents (or, in the case of the stage theory,
exdurents), the present system supplements endurents with processes. Some might regard this as an
affront to parsimony, but parsimony is only a factor when all other things are equal. Proceeding according
classical quantification theory and predicate logic has forced the rejection four-dimensionalism, and so all
other things are not equal. Hence, any impermanent objects must be defined in terms of permanent objects
instantiating properties at some time or span of time.
3. Being and Becoming: The Rationalist versus Empiricist Divide
One can understand why empiricists such as Quine, Lewis, and Sider would opt for four-dimensionalist
ontology and why staunch empiricists like Hume would understand the world as a world of events.
Indeed, those who tend to associate the meaning of names with descriptions will be inclined to identify
objects with lifetimes, and do away with talk of endurents altogether. Again, as far as ordinary language is
concerned, one will typically do well to understand the names we commonly use as referring to lifetimes,
and other processes, rather than substances. So it will be admitted that people in general are, indeed, more
interested in processes and lifetimes than they are in any underlying objects. But I never supposed that
ordinary folk have much interest in taking up the view from nowhere and, my experience is that ordinary
folk typically have no such interest.
The empirical world is truly one of events and processes; after all, we are never directly
acquainted with necessary objects as they are in themselves, and so the empiricists will tend believe that
the world is exhausted by events and processes. As Descartes recognized in the Second Meditation,
experience as such does not acquaint us with the notion of substance; the notion of an enduring substance

204

is added by thought. It could be that thinking, in adding substance, adds the lie to experience (to evoke
Nietzsche) or it could be that it reveals it.
Here is my broadly Platonic-Rationalist retort. In this context, I would read Platos allegory of the
Cave to indicate that, when we restrict our understanding of reality to what is presented by experience
alone (i.e., appearance), we treat reality as something like a shadow play a mere play of appearances.
The lesson, I take it, is that, upon reflection, we should come to judge that such appearances present
themselves as essentially dependent entities. Just as shadows do not present themselves as things with
causal powers, the world conceived wholly in terms of occurents presents itself as a world without
anything possessing power. Any recombination of shadow-shapes is possible (space permitting) precisely
because shadows are mere epiphenomena without natures of their own. The Humean-world is like the
shadow-play in the Cave but without the Cave, fire, or projected artifacts there are the shadows, and
thats all. Shadow-like objects without powers or essential nature are considered as if they could just exist
without being further explained. But the central Platonic-Rationalist judgment here is that it is precisely
this similarity with a mere shadow-play that justifies the inference that occurents could not possibly exist
all by themselves. An appearance must be the appearance of something and through something events
must be causally dependent upon the substances they involve.
Hume and his followers apparently do not form this judgment or, at any rate the temptation is
resisted. In Humes case this was the result of a deliberate empiricism that systematically avoided
explanations to the best explanation. In latter day Humeans it seems to be something like an acquired
taste. Humeans are quite satisfied coming up with theories about the regularities of shadow patterns on
the wall so as to explain how it is that some people are able to make good guesses about episodes that will
occur during future acts of the play. The prisoners in the cave, of course, appreciate this sort of
instrumental knowledge: knowledge focused on how we try to find our way around the world rather than
knowledge arising from the desire to know things simply as they are.
From a Platonic point of view, Humeans never even turn their attention to the artificial objects
and the fire, much less toward the light shining down from above. In Chapter One I interpreted the
turning, or conversion, emphasized by Socrates as the transition from desiring to know things for
conventional or pragmatic purposes to desiring to take up the view from nowhere which involves a
complete reorientation of the soul. When this reorientation has taken hold, one begins to become a
philosopher in the Platonic sense of the term; one comes to be interested in knowing Being more so than
Becoming, and Becoming only inasmuch it is an image or effect of being.
Df105: Surface Process: An event ee is a surface process, written SPee, if and only if:

205

Pee ^ !e(((e ee) ^ e) Me)
Surface processes include only those processes which do not include any fundamental property
instantiations. Such instantiations, being necessary and permanent, are indeed very far removed from our
ordinary usage of the term event and even further removed from experience. Sufrace process will not,
in general, be a causal process, at least if all causal power depends upon the fundamental properties. This
will be more readily understood if it is noted that surface processes are really abstractions from real
causal processes. For instance, while compositional and extrinsic natural properties are partly explained
by natural relations they are also partly explained by the natural properties of their relata. If those
properties should also turn out to be compositional or extrinsic then, by A3, we must continue our search
until we resolve all relational natural properties into fundamental natural relations and properties.
Df106. Core-Process: An event ee is a core process, written CPee, if and only if:
Pee ^ !e((e ee) (e ^ Me))
Between core processes and surface processes we have a partition all of processes into two classes.
Regarding the class of core processes, the tensing of the copula is truly superficial, since the involved
monadic properties are always necessarily possessed by their bearers. Given that fundamental property
instantiations are factually equivalent to the fundamental objects themselves, we may regard Being as
precisely the class of fundamental objects. In this way Being can be fully distinguished from what might
be regarded a proper processes (i.e., those involving change over time). Regarding the class of surface
processes, the tensing of the copula is typically informative, since they typically involved properties and
relations that are not always possessed by their bearers. The class of core processes represents the realm
of Being while the class of surface processes represents the realm of Becoming.
Being consists of those entities that can be thought but cannot be sensed while Becoming consists
of those processes which can be sensed but cannot be thought as existing in themselves. It is partly virtue
of being related to sense that Becoming is an abstraction from Reality and it is in virtue of being an
abstraction from Reality that it is a-causal. Of course, Being here is also an abstraction, since the
fundamental entities do bear temporary and contingent relations to one another, which relations result in
the contingent properties of surface process. The difference, however, is that in the case of surface
processes, the derivative has been abstracted from the fundamental while in the case of core processes the
fundamental has been abstracted from the derivative.
For Plato, Being was apparently a realm of unchanging Forms, which are typically taken to be
ante rem universals. The present theory is a sort of compromise between Plato and Aristotle (or perhaps
Plato and Parmenides). Forms do not seem to be the right sort of thing to play the role of fundamental

206

object but the realm of Being consists of necessary and permanent substances and their fundamental
properties, which are also necessary and permanent in that they are nothing over and above their bearers.
The realm of Becoming consists of contingent events and processes which are grounded in contingent
external relations between objects.
Being (Fundamental Domain): Becoming (Surface Processes):
Necessary and Permanent Existence Contingent and Temporary Occurrence
Necessary Fundamental Properties Contingent Extrinsic Properties
Elusive Properties Observational Properties
Transworld Identity Counterparts
This treatise began by taking up Platonic-Rationalistic methodology in applying the geometric method
based upon the fundamental principles of logic and being-qua-being, Platonic-Rationalistic content has
arisen as the final result.
4. What Might Play the Fundamental Substance Role?
As for location problems, the primary question does not concern how various roles are played by
various parts or subsets of the fundamental base but, rather, which specific theoretical objects might end
up playing the fundamental base role. Given the fixity of existence, there can be no straightforward
identification of members of the fundamental base with anything that has been studied in physics and it is
unlikely that physics will ever come to empirically discover such objects. What then are the prospects of
physicalism? At most, it may turn out to be the case that each fundamental physical object is a lifetime
of some fundamental object and, at any given time, each fundamental object may manifest itself in terms
of some physical lifetime. The second conjunct, however, is challenged by the fact it seems that there
were far fewer physical objects in the distant past that there are now. This problem might be dissipated
should it be the case that we live in a multiverse in which the total number of physical objects is constant.
The result, then, is that, a revisionary form of physicalism remains on the table, but its status is quite
uncertain.
More precisely, something like physicalism could be true of all extrinsic properties are observable
properties (in which case the two classes are liable to be co-extensive). If this were so, then all genuine
happenings (i.e. all change) would be physical.
Where else could we turn when looking to find out which objects play the fundamental object
role if not the objects of some future fundamental physics? It is at this point that it becomes evident that

207

our knowledge of things themselves is bound to be quite limited, but this should not be terribly surprising.
At any rate, entities satisfying some of the characteristics of necessary substances have been put forward
in various different contexts and it might be worth our time to briefly take note of these.
Within the tradition of rationalistic philosophy, necessary substances bare some similarity with
Leibnizian monads in that they persist throughout many lifetimes or transformations and in that they are
not what one usually has in mind when speaking of physical objects. My theory of necessary substances,
however, differs from Leibnizian monadology in important ways.
First, Leibnizs metaphysics is a kind of idealism (the monads are intrinsically mental) while the
present system is perfectly neutral regarding the question of panpsychism.
Second, monads have all of their properties essentially. Essentially here does not entail
permanence, monads alter over time, but each such alteration is part of its sequential essence. In this
respect monads resemble lifetimes in that the only thing contingent about them is whether or not they
obtain. The fact that monads are supposed to be substances, however, means that, given NE, if the
necessary substances were monads, modal collapse would ensure, in violation of A1.
Third, and related to the second, monads are said to be powerful but there is no real causal
interaction between them. Hence monadology may be true to the letter of A2 but it is not true to its spirit.
There is a second sort of panpsychism with some affinity with the results of this study, namely
certain traditions inspired by the metaphysics of the Upanishads. If there are necessary objects underlying
the lifetimes of organisms such as ourselves (i.e., if each I refers to a member of D) then most of what
we typically regard about ourselves pertains not so much to ourselves as substances but as surface
processes, and this includes not only bodily manifestation but also our empirical consciousness. In this
respect, the present theory is not only somewhat Kantian (although nothing has been established here
regarding anything like noumenal freedom), but also Hindu.
The Upanishads themselves seem to say that everything is identical with atman, being the self
of the universe and frequently identified with Brahman, or the Absolute. According to the Advaita
tradition, Brahman is literally the true self of each individual (i.e., all plurality is illusionary), the result
being an ontology that adheres to both pantheism and strict monism (i.e., we are not even really modes
of the absolute). While the Advaita interpretation certainly follows the letter of the Upanishads it is
representative of a minority of Hindus, no doubt in part because of the fact that all manifestation turns
out to be a veil or illusion and all desire is to be negated. According to the Vishishtadvaita tradition,
however, the identity of the real self with atman, and hence Brahman, is understood not in terms of
numerical identity but in that the inner-self of each creature is co-eternal with and shares in the same

208

general nature as atman or Brahman. There is a single cause and ground of all things, but the effects of
this cause are real, consisting in its own internal development with internal multiplication.
161
The results
of the present treatise suggests that the Advaita tradition would be right in claiming that atman is not only
eternal but (intrinsically) unchanging while the Vishishtadvaita tradition would be right in claiming that
atman is multiple and that identity with the Absolute is, at most, an identity in basic nature.
According to the Upanishads, suffering is the result of identifying oneself primarily with ones
empirical self or ego (ahamkara) rather than with ones atman, which is unchanging, eternal, and
internally related to all other things. Indeed, in many ways Humeanism is the antithesis of Hinduism.
Hume identifies the self directly with the bundle of perceptions one has at any given moment and saw
the world as a collection of disconnected happenings.
The atman, however, is thought to possess a certain kind of consciousness and, as mentioned
above, this treatise has provided little reason to conclude that we, or any other being, is intrinsically and
perpetually conscious, even in some broad sense of the term. It is worth noting, however, that if one were
to combine the conclusions of this treatise with the sort of panpsychism defended by Strawson (2006),
then one would arrive at a metaphysics that produces many of the core elements of Hindu metaphysics.
Making such a claim, however, is a strong move. It could, apparently, only be motivated if (1) one
acknowledges Cartesian (or rather, Augustinian) certainty of ones own mental existence and, (2) judge
that the mental cannot be identical with or emerge from an collection of entities, each of which is
intrinsically non-mental (i.e., mental properties of some kind or other must be fundamental). One might
also make use of Berkleys argument that we have no real concept of a mind independent substance,
given that our only examples are qualities seem to be derived from qualia.
While I believe that exploring this avenue might be worthwhile, I also perceive a tendency among
some theorists to jump from non-physical to mental, as if physical and mental were logical
compliments of one another. I think we should take seriously Spinozas claim that extension (i.e.
physical) and mental represent but two of the infinite attributes of substance. Perhaps infinity here is
unnecessary, but the point is that, our ignorance of strongly intrinsic properties is just that, a form of
ignorance. Given our ignorance the most likely thing to conclude, all other things being equal, is that the
fundamental properties are neither physical nor mental. That is, we might do well to adopt a kind of
neutral property monism whereby there is but a single kind of substance which is neither fundamentally
physical nor fundamentally mental. The drawback of neutral monism is that it does nothing to flesh out
our understanding of fundamental reality beyond the necessary conditions required for something to play

161
Co-eternity with Brahman is retained in virtue of the doctrine that the effect pre-exists in its cause
(Satkaryavada).

209

the fundamental object role, that is, it does not really take us beyond the basic theory of substance
outlined in this text. We would like to be able to do better, but we must first, and carefully, consider
whether or not we can.
Finally, on the topic of Hindu philosophy, my repeated insistence that, when doing metaphysics,
we should do what we can to take up the view from nowhere might be understood to mean that we seek to
identify ourselves with something like atman, or how we are with respect to Being, rather than with
ahamkara, or how we are with respect to Becoming. Accordingly the metaphysical injunction, which is
directly motivated by the desire to know Reality in itself, might at once be a spiritual or mystical
injunction to strive, as far as we can, to become one with the cosmos. It should also be expected that this
practice will result in a sound ethical disposition, whereby one will not be inclined to judge ones own
cases more favorably than one would judge the case of another.
5. A Note on Locating Manifest Substances
I began this treatise by suggesting that one should develop a theory of Reality in itself before addressing
location problems, since the latter really does presupposes the former. I complete this treatise by briefly
considering how one might approach location problems in light of our findings. What I will have to say
here is mostly quite general in that the following considerations should be evoked prior to the
investigation of any location problem.
(1) Can a lifetime of a substance include parts of the lifetimes of other substances (i.e., are
composite substances possible)? If so, which natural kinds count as such lifetimes?
(2) Are we justified, via Cartesian-Augustinian certainty, in taking ourselves to be
substances? If so, does this justify some form of dualism or only the claim that our bodies
should count as substances?
(3) Are we to regard simple particles as individual objects or as mere excitations of global
physical fields?
The first question concerns the matter of material composition restricted to the question of composite
substances or, more accurately, whether a lifetime can include segments of other lifetimes. Coming to
terms with (1) would seem to require finding some explanation as to how the lifetime of one object can
take up the lifetimes of various other objects or, perhaps, give rise to other, dependent lifetimes. Ones
approach to each of these considerations will be interdependent with ones approach to the others. If
certainty of oneself as a substance justifies belief that ones body is a substance then one might extend
this belief to other humans, animals, and plants. If elementary particles are lifetimes of substances then

210

they would have to be taken up into the lifetimes of other objects, such as molecules or cells. This
presents difficulties, since the molecules and lifetimes seem to begin with the coming together of their
elements and so it seems wrong to say that the whole is taking up its parts into a process precisely because
the lifetimes of the elements are manifestly independent of the lifetime of the alleged composite. That is,
if one begins with a micro-reductivist view, it is difficult to fathom how there could be genuine composite
lifetimes.
The situation might be easier if we take point particles as mere excitations of global fields
whereby the electromagnetic field, for example, would be a lifetime of a substance but no individual
photon would be a lifetime of anything.
162
In this case the simplest spatially local lifetimes might be
organic cells. In which case individual cells can be treated as mereological simples (at least in that they
never have a segment of a lifetime as a proper part) and hence an organic lifetime would begin as a simple
lifetime, such as a zygote. Any lifetime that comes to be a part of such a lifetime will come to be a
lifetime in terms of growth from the initial zygote. Indeed, we would do well to keep in mind the
difference between growth and aggregation. Growth occurs in virtue internal genetic mechanisms and by
way of division, while aggregation occurs by way of addition. In cases of division the whole is prior to its
parts
163
while in cases of aggregation the parts are prior to the so-called whole. To be sure, growth
requires nutrients, but by hypothesis, the nutrients in questions are not lifetimes of any substances. When
absorbing a molecule a cell is simply interacting with global fields in certain ways. This, of course, would
suggest that living cells are something over and above global field interactions, and hence exhibit some
sort of ontological emergence (i.e., its interaction with the global fields is not just the global fields
interacting among themselves). It would also require that nothing that can be absorbed by a living cell is
itself a true composite unless it was produced within, and secreted from, another cell. There are, of
course, other ways in which one might seek to find genuine or substantive wholes, some of which might
not involve such a bio-centric viewpoint. I simply wanted to conclude by suggesting a general approach
that one might take regarding the question of composite substance.



Copyright Adam Labecki 2013

162
The same would hold for fermions, such as electrons, which would be excitations of the electron field.
163
Although it must be admitted that the initial divisions of the zygote has the appearance of a mere aggregate of
cells (the morula) bunched up in a ball. After a day or two the formation of desmosomes and gap junctions, both
bind the cells together and permit sub-cellular material to pass between the various cells.

211

APPENDIX A:
WHY I AM NOT A HUMEAN

As noted in Chapter One, metaphysicians within the neo-Humean tradition, led by Quine, Jackson,
Lewis, and Sider, do not have much to say about the details of fundamental reality. This is partly because
the details of fundamental reality are outsourced to physics and partly because much of their work
consists in providing reductivist accounts of the postulates of ordinary language. It will be seen that the
Humean bases defended by these theorists do not mesh well with what we do know about fundamental
physics and it will be argued that the whole business of locating ordinary talk relative to the base is
effectively a red herring that distracts from the truly radical, and in my judgment unacceptable, nature of
Humean metaphysics.
The desire to explain all that is typically deemed to be true in ordinary discourse in terms of
fundamental physics leads one to misrepresent both. On the one hand, one must represent fundamental
physics as if it were analogous the manifest image, which can only be done if the world is fundamentally
classical. On the other hand, one must represent the meaning of ordinary language and concepts as if it
depended upon whatever is taking place in the microcosm. In what follows I shall only defend the first of
these two claims, that Neo-Humeans do continue to treat physics as if it were classical and that they must
do so, given their overall theoretical commitments.
1. Humean Supervenience
While most metaphysicians are self-styled naturalists and like to think of themselves as among the
scientifically enlightened, most do not engage much in physics or even the philosophy of physics. In
Chapter One it was noted that the present state of physics does not really prescribe a particular
fundamental physical ontology. One response would have been to become engaged in evaluating the
various options that are on the table and to constrain ones theorizing to theories that at least match up
with the points on which there is a general consensus. This is not what has happened in practice. Instead
of engaging actual physics most mainstream metaphysicians have recourse to toy model theories which
are supposed to be neutral with respect to any way in which actual physics may ultimately fill out the
details. The problem, however, is that these toy models are almost never neutral; in particular, they
typically require that the world should turn out classical (local and wholly determinate or separable) and
made up of a large number of little objects. When introducing such toy models, metaphysicians typically
claim that they are making use of a classical model for the sake of simplicity which suggests that one
could just as well make use of a non-classical model. But this is misleading. Non-classical physics is not

212

just classical physics with the addition of some non-essential complications. Moreover, non-classical
physics admits of non-distortive simplifications of its own. Why not use those? Because such models
would not be illustrative of the picture of reality that metaphysics typically have in mind they do not
permit the world to be a Humean world, which must be local and separable: there can be no non-
spatiotemporal natural relations between things and everything is wholly determinate in complete
independence of the rest of the universe.
Even though QM is bound to be superseded by some more fundamental theory, Bell (1963)
demonstrated that any empirically adequate successor of QM will be either nonlocal or fail to satisfy
counterfactual dependence (CFD).
164
Indeed, the more mainstream interpretations, such as the
Copenhagen and GRW, satisfy neither locality nor CFD. This practically guarantees that the world is not
classical in a way that is acceptable to Humeanism unless some version of the Many Worlds
Interpretation (MWI) is true
165
. As we shall see, however, WMI poses distinct problems for fans of
mereologically based classical toy models.
For a well-known toy model of fundamental reality consider Lewis doctrine of Humean
supervenience:
Humean supervenience is named in honor of the great denier of necessary connections. It is the
doctrine that all there is to the world is a vast mosaic of local matters of particular fact, just one
little thing and then another. We have geometry: a system of external relations of
spatiotemporal distance between points. Maybe points of spacetime itself, maybe point-sized bits
of matter or ther or fields, maybe both. And at those points we have local qualities: perfectly
natural intrinsic properties which need nothing bigger than a point at which to be instantiated. For
short: we have an arrangement of qualities. And that is all. There is no difference without
difference in the arrangement of qualities. All else supervenes on that (Lewis 1986b: ix-x).
It should be noted that Lewis regards most of his work as a prolonged campaign on behalf of Humean
supervenience. This doctrine lies at the core of his metaphysics. It is the best theory of the world he
seeks to defend. The theory is not equivalent to any physical theory, although it is evidently inspired by
classical physics in general and Special Relativity in particular. The theory cannot be defended by
becoming more developed and detailed because, again, natural science is supposed to discover the details
about the natural qualities. Hence the whole campaign will be to show how we can locate ordinary
objects, events, identity through time, counterfactuals, laws, modal language, and the mind, within such a
manifold. The motivation for the project, it seems, is twofold. First, on a general level, Lewis, like many
other naturalists, wants to defend physicalism. The Humean base is supposed to be wholly physical and so
the project of showing that everything we need to talk about supervenes on such a base is a physicalist
project. Second, Lewis is defending Humean supervenience as a specific physicalistic way of conceiving

164
These will both be explained below.
165
As noted earlier, on MWI, the failure of CFD does not seem to produce a non-classical world.

213

the fundamental physical level. When taking the view from nowhere, however, all that there is the
Humean base. Such a desert landscape seems to lack all of the flora and fauna one might have taken to be
present in the world. The project, then, is to connect such an austere world view to our ordinary
pedestrian view of the world; to make it a world in which we can find ourselves and make sense of our
ordinary beliefs.
Let us call the manifold of propertied points the Humean base. For simplicity, let us assume
that spacetime itself instantiates the perfectly natural properties (i.e., let us assume supersubstantialism,
which Lewis otherwise endorses). Indeed, the base would be less Humean without such
spatiotemporalism. Hume, in keeping as close as possible to bare experience, perceived the world not as a
world of objects but a world of events. If all that there is in the world are propertied spacetime points,
there are no endurents, only occurents (temporally extended sets of which are perdurants).
The matters of fact comprising the Humean base must be as lose and disconnected as Hume
imagined events to be. That is, the spacetime points must be wholly independent of one another like the
frames on a film. The frames happen to follow a certain order. In most films the order will involve
phenomenological regularity and narrative coherence; but there is nothing to stop one from splicing the
film into bits and then taping the frames back together in any sequence whatsoever. Moreover, there is
nothing to prevent one from splicing each frame into pieces and taping them back together in any way
that fits the width of the reel. That is, the Humean base is radically local; each point exists and is the way
that it is independently of every other point.
In order to be perfectly local, the natural properties are intended to be wholly categorical or
occurent, which is to say, no dispositional or powerful properties are admitted into the base. No property
points beyond itself, as it were; every property is indicative of a way something is (i.e., a quality) but
nothing about the way that it, or anything else, could be. In order to ground laws based on regularities, a
four-dimensional manifold will be needed: it includes all past and future spacetime points. Given the
absolute independence of each spacetime point, no spacetime point is in any way dependent upon the
goings-on at earlier spacetime points. Each spacetime point is what it is wholly in virtue of itself.
In order to ground modal claims, a four-dimensional manifold is still not enough, so Lewis also
posits an infinite plurality of such distinct manifolds, one for every possible configuration of propertied
spacetime points involving both actual properties as well as properties which are completely alien to the

214

manifold in which we ourselves exist (i.e., the actual world).
166
Since a single desert landscape will not
do, a vast plurality of desert landscapes must be posited.
167

2. Humean Supervenience is Incompatible with Single-World QM with Collapse
Mauldin (2007: 50-61), Ladyman and Ross (2007: 148-151), Karakostas (2008) and Darby (2012) take
Lewis to task regarding the locality requirement of Humean supervenience. As we shall see, the locality
in question is not precisely the sort of locality described in Bell (1963) but I shall begin by noting that sort
of non-locality.
A key feature of QM is that the Schrdinger equation for the evolution of the wavefunction of a
physical system frequently predicts two or more mutually incompatible states of the system. This has led
to a probabilistic interpretation of the equation and this assumption has matched the statistical distribution
of experimental observations remarkably well (although it will still be noted that the Schrdinger equation
is not explicitly a probability equation).
On the the traditional Copenhagen interpretation of QM, an object in a superposition with respect
to observable O does not possess any determinate O-value until such time that it is measured. Upon
measurement the superposition is said to collapse into one of its component states. This feature of the
theory has led to what is known as the measurement problem: what feature of measurement-events
should be assigned responsibility for the collapse? There have been several theories regarding how
measurement could induce collapse and they can generally be divided into those which point to contact
with a larger physical system, i.e., the measurement device (Heisenberg 1958), and those which point to
the conscious observer as such (Wigner 1961).
168
Some collapse theories try to avoid the measurement
problem altogether by suggesting that it happens independently of measurements and observations.
Perhaps measurement events belong to a larger (less observer-centric) class of events which tend to bring
about collapse. Perhaps collapse occurs spontaneously and frequently enough to prevent us from ever
observing superimposed macro-states (the Ghirardi-Rimini-Weber (GRW) interpretation).
While collapse theories, of one kind or another, are the norm, there are also non-collapse
interpretations of QM. The most prominent theory of this kind is the Many Worlds Interpretation (MWI)
(Everett 1956/1973, 1957), according to which superpositions are understood to describe definite states

166
Lewis does permit worlds that might have non-spatiotemporal external relations, emergent properties, etc., but he
considers these to lie within the outer sphere of possibility.
167
Those Humeans who reject this last part of the story are bound to hold deflationary views of modality. Quine was
such a deflationist for most of his career, while Sider (2011) has recently become a deflationist. I shall argue below
that Humeans cannot easily get away with deflationism regarding modality.
168
The hypothesis that conscious observation induced collapse was first proposed by Von Neumann.

215

which obtain at different possible worlds. MWI theorists often describe the world as branching
whenever the collapse theorists would describe an indeterminate state, but it is probably best take this
branching to be qualitative in nature rather than numerical. That is, one should regard all of the many
worlds as existing from the beginning (if there is a beginning of time) but that sometimes worlds w and w
are perfectly similar up to a given time t, at which point they begin to differ or branch.
Entangled states are superpositions of physical systems involving more than one particle.
Suppose that e
1
and e
2
are two electrons and have been entangled so that the wave function results in a
superposition of <e
1
having z-spin up and e
2
having z-spin down> and <e
1
having z spin down and e
2

having z-spin up>. This in itself already spells trouble for Humean supervenience. It is widely believed
that the information encoded in the wavefunction for a particle x with respect to an observable F is
complete; everything there is to be known about x with respect to F at time t is indicated in the by the
state of the wavefunction at time t. Assuming, then, that the wavefunction tells us everything about the z-
spin of the two particles, neither e
1
nor e
2
appears to have a definite spin. Spin number among the basic
properties of the Standard Model, which present-day Humeans typically identify with the fundamental
properties. Nevertheless, it seems that it happens that sometimes a particle can have indeterminate spin.
As Ney (2010: 227) notes, e
1
can be said to have the extrinsic property of having the opposite
spin as e
2
and, likewise, e
2
has the property of having the opposite spin as e
1
. Moreover, the relation in
question is not internal in that it is not one that rises from the intrinsic states of e
1
and e
2
(the
wavefunction does not assign any information that could ground such a relation) and so it must be
external. The only sort of external (or natural) relations permitted by Humean supervenience, however are
spatiotemporal relations, but the entanglement has nothing to do with the spatiotemporal relations
between e
1
and e
2
, since they can fly to opposite ends and the galaxy and the entanglement need not alter.
Hence, one must conclude with Butterfield (1992) in affirming entanglement relations are both external
and non-spatiotemporal. Moreover, particles which are entangled with respect to all of their quantum
properties will not have any intrinsic quantum properties at all. This all spells bad news for Humean
supervenience, since it seems that particles do not always have a full list of wholly determinate physical
(i.e., observable) properties.
As for the famous violation of locality, on the traditional Copenhagen interpretation of QM, if x
and y are entangled with respect to F, then measuring x for F will result in a collapse of the entangled
state such that both x and y will come to be characterized by their own individual wavefunctions. In terms
of the example above, e
1
will be found to have either z-spin up or z-spin down. If it should have z-spin
down then we know that e
2
must have z-spin up (and vice versa). The non-locality at work is that a
measurement on e
1
will not only cause e
2
to obtain a determinate spin but it will do so instantaneously,

216

even if e
2
is many light years away. There are, of course, problems with the thought that measurements
alone initiate the collapse of the wavefunction and this is widely regarded as the greatest shortcoming of
the Copenhagen interpretation of QM; but whatever it is that causes the collapse will be an event that
need only locally act upon one member of an entangled system of particles and hence every collapse
theory turns out to violate locality.
169

It is not this sort of non-locality that does the real damage to Humean supervenience: the
reductivist about causation does not even believe in the local action of one event upon another much less
genuine action at a distance. Not only is there no action at a distance for the Humean, there is no action
at all. As Mauldin (2007: 61) points out, it is really the entanglement of states in general that do the
damage. When e
1
and e
2
are entangled with respect to z-spin, the quantum system, which includes them,
does have a determinate z-spin value, namely a spin of zero. Since neither electron has a determinate local
z-spin state, the quantum state attributed to the system cannot supervene upon its mereological
components and hence Humean supervenience fails.
170
That is to say, the progression of the wavefunction
does not supervene upon the Humean manifold of a given world.
Just to recap. On mind-independent one-world QM theories with collapse:
(1) Objects do not always have a complete set of determinate observable local properties.
(2) The facts encoded by the wavefunction require non-local external relations (i.e.
primitive entanglement relations).
(3) Facts about collections of quantum states often do not supervene upon the members of
a Humean base plus the spatiotemporal relations that obtain between them.
From these considerations it is very clear that Humean supervenience cannot be true if any one-world
interpretation of QM with collapse is true.
3. Humean Supervenience can only be Trivially True According to MWI
One might suppose that defenders of Humean supervenience (or something near enough) might
seek refuge in the MWI of QM. Aside from positing many parallel worlds, WMI is effectively classical.

169
Or rather, the Relational interpretation (Carlo Rovelli 1996) is the one collapse interpretation on the table that is
also local, but it is also non-objective in that the state of any quantum system is always relative to the reference
frame of the observer (i.e., one could not speak of a mind-independent physical base). Moreover, despite the
satisfaction of locality (in Bells sense of the term), on the Relational interpretation, as the name indicates, QM only
tells us about how physical systems relate to other physical systems and such relational knowledge is thought to
provide a complete description of the world.
170
The same case against Humean supervenience is raised by Ladyman and Ross (2007: 150). Also, see Teller
(1986) for a very holistic interpretation of entanglement as a form of emergence.

217

Moreover, if one is already willing to embrace Modal Realism, little harm is done if one also accepts the
worlds of QMI. There is, however, one catch: it seems to require that one adopt existence monism which
will not be welcomed by Humeans:
Since the universal validity of the state function description is asserted, one can regard the state
functions themselves as the fundamental entities, and one can even consider the state function of
the entire universe. In this sense this theory can be called the theory of the universal wave
function, since all of physics is presumed to follow from this function alone (Everett 1956/1973:
8-9).
Not only does the Everett interpretation adhere to a form of wave-function realism (what Everett is here
calling the state function), it makes good use of Ockhams Razor in making it the only fundamental
entity, where the plurality of worlds are merely its different states: there are infinitely many worlds but
only one fundamental and trans-world object! This, in and of itself, does not quite require full-fledged
existence monism, only what Schaffer (2009) calls priority monism (there is but one fundamental object).
It might be the case that WMI is ultimately even committed to existence monism (the doctrine that there
is but a single physical object), but since the Humean base is supposed to be fundamental, I shall only
explain why it leads to priority monism. The following argument might be parsed-out of the reasoning
implicit in the Everett passage cited above in conjunction with Neys (2010) argument for wavefunction
realism:
1. Wavefunctions physically exists (Wavefunction Realism)
2. Information encoded in wavefunctions exhaust the empirical content of QM and hence
provides complete information about quantum states
(Wavefunction completeness)
3. There is no need to posit any fundamental physical objects besides wavefuntions
(regardless of how wavefunction realism is to be understood) (1-2, from parsimony)
4. The Universal wavefunction is the only wavefunction (No Collapse)
5. The Universal wavefunction is the only physical object (3-4)
Commitment to MWI involves commitment to (1), (2), and (4), where the first two are commonly held by
physicists and where the fourth is characteristic of no-collapse theories. Regarding (4), No Collapse
requires that systems never cease to be entangled but systems become entangled very easily, and so even
if there was a past time of many independent systems (which is unlikely given the Big Bang theory) it can
be expected that the entire universe is entangled by this point. From (1) and (2), (3) seems to be a pretty
straightforward application of parsimony. On MWI the entire multi-verse at any time is one great big
superposition of states, each one corresponding to a possible world. Given WMI, the only reason the

218

whole can supervene upon the parts, is because the only part of the multi-verse is the whole darn thing,
whose state vacuously supervenes upon it.
Humean supervenience is effectively incompatible with any theory that entertains superposition
of states (including entanglements). Superposition will lead to the violation of Humean supervenience
unless there is only a single object, but this still violate the whole spirit of Humeanism (since the
individual objects of experience are not independent entities susceptible to recombination) and renders the
location problem intractable (in the least one will be unable to locate ordinary objects, laws, etc. with sets,
pluralities, or sums, of fundamental objects). If this is correct, Humean supervenience is incompatible
with any interpretation QM or any current theory of quantum gravity (all of which involve superpositions
and would be empirically inadequate otherwise).
171
In short, Humean supervenience is no more scientific
than Scientology and, given the fact that it is a theory of the physical world, it is irrational to suppose that
it is true.
While Humean supervenience is not the only toy-model of fundamental reality employed by
mainstream metaphysicians, it is typical of the toy models one encounters. Armstrong (1997), for
instance, advocates the following view:
The spacetime world is a structured (that is, related) set of spacetime points. These points are the
fundamental particulars. That the points have certain properties and are related to each other in
certain ways constitutes the fundamental states of affairs (Armstrong 1997: 6).
Indeed, a common feature of these theories is what L.A. Paul (2012) calls spatiotemporalism, basically
the idea that the fundamental building blocks are propertied spacetime points (i.e. supersubstantialism).
Sider (2001, 2011) endorses the same sort of ontology. While the above passage from Armstrong does not
preclude the possibility of non-spatiotemporal relations, he argues for an independence principle that
requires each distinct object to be wholly independent of each other distinct object. Moreover, given the
fact that GR is an effective theory, spatiotemporalism is a suspicious doctrine. Indeed, Armstrong himself
is well aware that many proposed successor theories will construe spacetime to be phenomenal or
emergent (Armstrong 1983).
If the world is not fundamentally classical, the prospects for reconciling the manifest image with
the scientific image are slim. Progress within location problem (i.e., functionalists) metaphysics is not
of the sort found in the natural and formal sciences. Rather, as a reconciliation project, any such
progress is very much Scholastic. The number of reductive analyses may increase. Total theories might

171
Hope that physics will ultimately vindicate Humean supervenience is as misguided as the hope that it will
ultimately vindicate Aristotelian physics or the I Ching.

219

become more elegant. But if it is all based on a vision of fundamental physical reality that is inconsistent
with our best present and anticipated physical theories, then the whole project is doomed from the start.
4. Philosophical Objections to the Metaphysics of the Humean Base: Fundamental Reality
as Pure and Utter Chaos
Lewis defense of the tenability if Humean supervenience consists in showing how one can define various
concepts and classes of objects in terms of the Humean base, especially when it has been sufficiently
multiplied. I do not think that such a line of defense should be convincing, at least not to anyone who is
not already committed to Humeanism. If one wants to combine Humeanism with systematic metaphysics,
then Lewis certainly shows the way one should go but unless one is already willing to accept the Humean
picture of the world, no functionalist analysis in terms of supervenience can provide a good reason to
change ones mind.
Let us turn to the Human picture of the world while taking up the view from nowhere, without the
mediation of any functionally defined terms. Let us take a look at just what sort of theory of fundamental
reality is at work here. Doing so is not distortive, since the functionally defined terms are precisely
defined in terms of the fundamental base anyway. Inasmuch as one attains the view from nowhere one
sees the world strictly in terms of its fundamental objects and relations. That is all that there really is and
hence, from a metaphysical point of view, it is all that really matters.
As soon as we focus our attention in this way, it immediately becomes evident that Humeans
effectively believe that something can come from nothing. There is just one little thing and then
another; the only fundamental external relations between the objects or events are spatiotemporal.
Fundamentally speaking, things just happen. There might happen to be some patterns to be found in what
just happens but it is purely coincidental. These patterns are of practical interest to us, but any theory of
regularities reveals nothing interesting about the nature of fundamental reality. In particular, there is no
causation in the sense of production or generation and so in this sense of the term, causation is not
revised so much as eliminated.
Given the non-fundamentality of causation or natural laws, each and every spacetime point
throughout the entire four-dimensional manifold of each and every possible world is as perfectly
fundamental as it would be if it marked the very beginning of time and was not even brought into being
by some eternal creator like God. Each and every contingent fact is a fundamental datum in that it is

220

simply given.
172
That is, Humeans systematically deny the traditional metaphysical axiom ex nihilo nihil
fit. Indeed, something coming from nothing is completely ubiquitous: every single spacetime point is an
instance of something coming from nothing. Each point is caused by nothing and the cause of nothing.
Moreover, anyone who holds to such a theory of fundamental reality ought to affirm that it is
impossible for anything to come from anything else. To be sure, Lewis (1986b) purports to defend
Humean supervenience only as a contingent and empirically based truth, which may not hold in the outer
sphere of metaphysical possibility. That is to say, he does not believe that the concepts of productive
causation, numerical identity through time, or non-spatiotemporal external relations are incoherent. It is,
however, surely not an empirical matter that these sorts of entities do not exist in our world. In light of the
previous section I think it is safe to say that we do not learn any of this from natural science since natural
science provides plenty of reason to reject the whole Humean picture. And surely, if such real causation,
numerical identity through time, etc., is possible at all, there are worlds epistemically just like ours which
contain such things. What justification would one have for thinking that such rubbish does not exist in
this world? Indeed, if one believes that such things exist in other concrete worlds, whence the judgment
that they are rubbish or metaphysical suspicious at all?
Accordingly, the only good reason one might have for believing that no such things exist (in the
actual world) is to believe that they are intrinsically incoherent. Indeed, I believe that Humeans, or the
full blooded ones at any rate, in their heart of hearts, do believe that the whole idea of production
(generation or creation) is so mysterious and baffling that we ought to take it to be incoherent.
While it may be possible, and perhaps an interesting exercise, to suppose that everything comes
from nothing and nothing comes from anything are true principles, one is unsure why one would
accept such claims as a starting point unless one was a radical empiricist who refused to make any
explanatory posits at all beyond what is given in immediate experience. Radical empiricists will also be
skeptics about the external world, other minds, etc. Neo-Humeans are not skeptics of this sort. Jessica
Wilson (forthcoming) suggests Humes denial of necessary connections retains currency only because it
permits one to hold recombination principle of modality and that it will continue to retain currency unless
an adequate replacement should be found. In Chapter Eight I proposed such a replacement.
Is ex nihilo nihil fit a fallible principle? Sure, but only because every principle is fallible. Do
principles come with greater certainty than ex nihilo nihil fit? Maybe a couple of logical and mathematical

172
One odd feature of Neo-Humean metaphysics is that it appeals to contingent givens (local matters of fact) which
are not empirically given. Even in the absence of possible worlds, at least half of the Humean manifold is in the
future, which cannot be observed in any way whatsoever.

221

axioms are held with greater certainty, but as far as metaphysical principles go, they do not seem get any
more certain or more basic.
Note that Lewis plurality of worlds is not only an interesting way to provide truthmakers to all
forms of modality, it is practically necessary if Humeanism is to make any sense at all. Things just
happen and anything can follow upon anything. Reality is literally chaotic and yet, among all the changes
we observe, there is a fair bit of continuity. While I did not exist a moment ago, someone very much like
me, my past self, did exist a moment ago and I have inherited nearly all of that persons qualitative
features, and a chain of such persons stretches back over 35 years. And Im not the only one! As far as I
can tell, and Lewis would agree, there are billions of such gen-identity worms on this little planet alone.
Yet, in each case, nothing about the preceding person necessitated or even inclined nature to make the
next region of spacetime hold such continuities. Things just happen. Anything can happen anywhere
(space provided) and all this regularity obtains by sheer coincidence!
Neo-Humeans, then, really must be modal realists. Every permutation of distinct property
instantiations over the four-dimensional manifold exists (including properties which are alien to this
world), and the vast majority of worlds surely present the sort of utter disorder that one would naturally
expect if anything can follow upon anything. But if you have enough chaos there will be pockets of
apparent order. Some worlds will be mostly chaotic but include the odd bubbling of law-like regularities.
Our world is one of the few that exhibits law-like regularities permitting both the formation of matter (as
we know it) as well as life.
173
On the Plurality of Worlds does not include an argument from
Humeanism. I believe this is because Lewis knows that red herrings about the usefulness of possible
worlds will be more persuasive. It is strategically more sound and persuasive to emphasize the
theoretical benefits of having a unified theory of modality, which serves as a basis of a total theory
including not only a theory of fundamental reality (the Humean base) but also accounts for ordinary
language in terms of the fundamental base. But what good is all of these ways of tracking our statements
in terms of the base if the base is most likely inconsistent with our best physics and has it that each thing
literally comes from nothing and that reality is, at bottom, completely chaotic? Considerations of ones
fundamental theory should always trump those of what Lewis calls total theory, which already
presupposes a Humean framework of theory evaluation (i.e. preference for reductive analysis is already a
Humean preference).
There is more. As I type this chapter (or rather, as a series of persons with qualitative similarities
type this chapter), I perceive a netbook with its display screen and keyboard. But I do not do so because

173
Although, we cannot be sure that the entire world is like this, since we can only observe a relatively small portion
of the entire four-dimensional manifold.

222

of any influence by which the netbook has some effect on my sense organs and from my sense organs to
my brain. Despite my present environment and state of my body, I could just as easily have perceived
myself to be taking a walk in Medieval Baghdad or floating around the Astral Plane. There is no reason
why my brain should produce perceptions that have anything at all to do with my immediate environment
and yet it seems that, at least when I am awake, it generally does. Again, this appears to be the case with
nearly all sentient beings on this Earth. Everything happens as if there were some pre-established
harmony of the sort one finds in Leibniz, only there is no God to ensure that everything harmoniously fits
together. Occasionalism and Berkeleyan idealism are certainly fantastic sounding theories that strike one
as very unlikely, but at least they are theories which offer some explanation for the phenomena. As
mentioned in Chapter Seven, once we rid ourselves of genuine causal relations realism about the external
world (not to mention other minds) becomes unjustified. Our perceptions of external objects (for identity
theorists like Lewis this means our brain states) are completely independent of what occurs in our
environment and hence they are in no way grounded in anything occurring in our environment.
Accordingly, there is no need to posit an external world in the first place (including brains). That is,
skepticism regarding causation is part and parcel with radical empiricism. One must either embrace the
entire package or reject the entire package. I, for one, chose to reject it.
The ontology behind Humean supervenience and the regularity theory of laws reminds one of the
shadow plays depicted in the allegory of the Cave. Socrates asks Glaucon the following question
regarding the philosopher who has returned to the darkness of the Cave, having just witnessed the
intelligible reality outside:
And if there had been honors and commendations among them which they bestowed on one
another and prizes for the man who is quickest to make out the shadows as they pass and best able
to remember their customary precedences, sequences, and coexistences, and so most successful in
guessing at which was to come, do you think he would envy and emulate those who were honored
by these prisoners and lorded it among them, or that he would feel with Homer and greatly prefer
while living on Earth to be a serf of another, a landless man, and endure anything rather than opine
with them and life that life?
Yes, he said, I think that he would choose to endure anything rather than live such a life.
(Republic 516c-d)
I think that is as good a place as any to rest my case against Humeanism and Humean supervenience.




223

APPENXIX B:
SKETCH OF THE OVERALL LOGIC

For the convince of the reader, this appendix contains a full sketch of the elements of the logic
employed in this treatise (minus the logic of branching world-stages), as well as the metaphysical proper
axioms.
Syntax
Lexicon
Primitive Terms:
5. Object constants: a
1
, a
2
. (a, b, c)
6. Object variables: x
1
, x
2
... (x, y, z)
7. n-place predicates constants (n 0): F
n
1
,

F
n
2
. (F
n
, G
n
, H
n
)
8. n-place predicate variables (n 0): X
n
1
,

X
n
2
. (X
n
, Y
n
, Z
n
)
9. Time Variables: t
1
, t
2
.
10. Singular Event variables: e
1
, e
2
.
11. Plural Event variables: ee
1
, ee
2
.
Primitive Connectives, Operators, and Logical Connectives
1. Connectives: , , , <, =, ,
2. Operators: !, .
3. Logical Predicates: =
4. Punctuation: (, )
Defined Connectives, Operators, and Logical Predicates
1. ^ is short for ( ) (Conjunction)
2. \ is short for (Disjunction)
3. is short for (( ) ( )) (Material Biconditional)
4. v is short for !v (Existential Quantifier)
5. is short for . (Possibility)
6. oo is short for o=o (Non-identity)
7. E!o is short for =o (Existence)
8. e is short for ((e = r
n
o
1
o
n
t) ^ (( = r
n
\ = o
1
o
n
))
174


174
Only a selection of the defined terms for the logic of events is included here.

224

9. (ee
1
ee
2
) is short for !e(e ee
1
e ee
2
)
10. St
1
t
n
is short for !t
i
(t
i
< t
i
< t
n
(t
i
t
1
t
n
))
11. ee is short for e((e ee) ^ e)
12. ee[] is short for !e((e ee) ^ e)
13. Eee is short for t(ee[t])
14. Pee is short for t
1
t
n
(St
1
t
n
^ !e((e ee) (e[t
1
]\ e[t
n
])))
15. is short for ( ( ^ )) ^ ( )
16. is short for ( ) ^ (( ^ ) )
1.2: Grammar
The Greek letters , , and are used as meta-variables denoting any formula of the language , r
n
is a
meta-variable for any n-place predicate, , , and are a meta-variables for terms, and v is a meta-
variable for variable terms in particular. The expression (v) denotes any WFF with at least one free
occurrence of v.
1. If r is an n-place predicate, where n 0,
1

n
are terms, and t is a time, then r
1

n
t is a
formula.
2. If and are terms then = is a formula, provided that both are object terms or both are
predicate terms.
4. If and are WFFs then , , and ., , are formulas.
5. If is any formula and v
1
... v
n
are variables, then !v
1
... v
n
is a formula.
6. If and are two event terms of the same number then = if a formula
7. If is a singular event and is a plural event, then is formula.
8. Only strings that can be shown to be WFFs using 1-7 are WFFs.
Logical Axioms and Rules
Propositional Logic
PL1: ( ) (Simplification)
PL2: ( ( )) (( ) ( )) (Frege's Axiom)
PL3: ( ) ( ) (Transposition)
MP: if and , then (Modus Ponens)
Uniform Substitution (US): The result of uniformly replacing any formula or formulas
1
, ,

n
in a theorem by any formula or formulas
1
, ,
n
is itself a theorem.

225

Classical Quantification Theory (CQT)
CQT1: !v /v (Universal Instantiation)
Provided that v and are both terms of the same type.
CQT2: !v( ) ( !v)
Provided that v is not free in .
UG: if , then !v (Universal Generalization)
Provided that v is not substituting for any term o that has been introduced by assumption.
The Logic of Identity
I1: = (Law of Identity)
I2: = ((o) ()) (Leibniz Law)
Axioms and Rules of S5 Modal Logic
K: .( ) (. .) (Kripkes Axiom)
T: .
S5: .
Modal Second Order Logic (MSOL)
MSOL1: X.!x
1
. !x
n
(Xx
1
. x
n
(x
1
. x
n
)) (Modal Second Order Schema)
Where X does not occur free in .
MSOL2: !X!(!x.(Xx x) X)
MSOL3: X
n
= Y
n
.!x(Xx Yx)
Temporal Logic:
t1: t < t (t-Irreflexivity)
t2: (t
1
< t
2
^ t
2
< t
3
) t
1
< t
3
(t-Transitivity)
t3: (t
1
< t
2
\ t
2
< t
1
) \ t
1
= t
2
(t-Linearity)
t4a: t
1
< t
2
t
3
(t
1
< t
3
^ t
4
t
1
< t
4
< t
3
) (t-Discreteness (a))
t4b: t
1
< t
2
t
3
(t
3
< t
2
^ t
4
t
3
< t
4
< t
2
) (t-Discreteness (b)
t5: t
1
< t
2
.t
1
< t
2
(Necessity of Priority)
Event Logic:
e1: !e(e) (ro
1
o
n
t/e) (e-CQT1)
e2: (e
1
e
2
) (e
1
= e
2
) (e-Identity)
ee1: !ee(ee) ((e
1
^e
n
)/ee) (ee-CQT1)
ee2: e(e) ee!e((e ee) (e)) (ee-Comprehension)

226

ee3: !eee(e ee) (ee-Non-Vacuity)
ee4: !e((e ee) (e ee)) (ee = ee) (ee-Identity)
ee4: (e ee) .(e ee) (Nec, of Containment)
ee5: ee !e((e ee) e) (ee-Occurrence)
Pee1: ee((ee ^ (pee ^ ee[]) ee((Pee ^ (ee ee)) ^ !e(e ee (e ^ e))))
(Pee-Comprehension)
Factual Equivalence
F1: ( ) (Identity)
F2: ( ( ^ )) (Idempotence)
F3: ( ^ ) (( ^ ) ( ^ )) (Commutativity 1)
F4: (( ) ( )) (Commutativity 2)
F5: ( ^ ( ^ )) (( ^ ( ^ )) (( ^ ) ^ )) (^-Associativity)
F6: ( ) (( ^ ) ( ^ )) (Additivity)
F7: (( ) ^ ( )) ( ) (Transitivity)
F8: ( ) .!t( ) (Necessary Logic Equivalence)
F9: ( ) .(( ^ ) ( )) (Fixity)
F10: ( ) ( ^ ) (Facticity)
F11: ( ) ( )
F12: ( ) (( ) ( ))
F13: ( ) (( ) ( ))
Ground Logic
G1: ( ) ( ) (Asymmetry of Full Ground)
G2: (( ) ^ ( )) ( ) (Transitivity of Full Ground)
G3: ( ) ( ^ ) (Facticity of Full Ground)
G4: (( ) ^ ( )) ( ( ^ )) (Full Ground Amalgamation)
G5: ( ) ( ) (F-G 1)
G6: ( ) (( ) ( )) (F-G 2)
G7: ( ) (( ) ( )) (G-G 3)
G8: ( ) ( ) (Asymmetry of Partial Ground)
G9: (( ) ^ ( )) ( ) (Transitivity of Partial Ground)
Proper Axioms
A1: !e(e .e) (Denial of Modal Collapse)

227

A2: .(X Y(Y ^ (Xxt Yxt))) (Thick Realism)
A3: .( ee((ee ^ ee) ^ ( ee))) (Fundamental)
A4: !X!x!t
1
((X ^ Xxt
1
) !t
2
D
(ccm)
(t
2
< t
1
(D
(ccm)
xt
2
^ (Xxt
1
m))))
A5: !D
(ccm)
!x!t(D
(ccm)
xt (m D
(ccm)
xt))
A6:SX:
S
X !x!t(Xxt (Xxt E!xt))
Semantics
Definition of a Model
A model, M, is M = (W, D, T, <, I) such that:
W is a non-empty set of possible worlds, with a distinguished member w
@
, designating the
actual world.
D is a non-empty set of objects: the domain
T is a non-empty set of times or moments
< is a relation on T satisfying the following conditions:
(1) T is a strict linear ordering:
(a) !t t < t
(b) !t
1
!t
2
!t
3
((t
1
< t
2
^ t
2
< t
3
) t
1
< t
3
)
(c) !t
1
!t
2
(t
1
< t
2
\ t
2
< t
1
\ t
1
= t
2
)
(2) For any t
1,
t
2
e T, if t
1
< t
2
, then there exists some t
3
e T such that t
1
< t
3
and there is
no t
4
e T such that t
1
< t
4
and t
4
< t
3
(i.e., T is discrete).
I is the interpretation function satisfying the following conditions:
(a) if o is an object constant then I(o) e D
(b) if r
n
is an n-place predicate constant then I(r
n
) assigns r
n
to a member to a set of n+2
tuples x
1
, ,x
n
, t, w
175

Let be a set such that, for each positive integer n, if r
n
is an n-place predicate, I (o
1
), , I(o
n
) e D
n
,
and t e T, then o
1
, , o
n
, r
n
, t e . is the set of all events, possible or otherwise.
Let V be a set of subsets of and for any world w and such that
w
, let [ ]
w
be a function assigning
to a member of V subject to the following conditions:
If S e []
w
and S e []
w
then {S, S} e []
w
For any
1
,
n
, {[1]
w
, , [
n
]
w
} = [
1
, ,
n
]
w

175
Alternatively, it could be said that I assigns each p to a function g: TW (D
n
).

228

Let be the variable assignment subject to the following conditions:
1. For each object variable x, [x] e D
2. For each n-place predicate variable X
n
, [X
n
] e D
n
W T
3. For each time variable t, [t]
e T
4. For each simple event variable e, [e] e
5. For each plural event variable ee, [ee] ,
The valuation-function V
M
for ML and a variable assignment , is a function that assigns 1 or 0 (true or
false) to each WFF relative to each w e W according to the following conditions:
[V=] V
M
(o =
t
) = 1 iff [o]
M
= []
M

[Vn] V
M
(r
1

n
t, w) = 1 iff [
1
]
M
, [
n
]
M,
[t]
M
e I (r)
(w)


[Vn[n] V
M
(r
1
2
r
1
1
) = 1 iff (I (r
1
)) , (I (r
2
))
[V] V
M
(, t, w ) = 1 iff V
M
(, t, w) = 0
[V-] V
M
( , t, w) = 1 iff V
M
(, t, w) = 0 or V
M
(, t, w) = 1
[V7
x
] V
M
(!x, t. w) = 1 iff for every o e D, V
M(o/x)
(, t, w) =1
[V7
F
] V
M
(!X, t, w) = 1 iff for every Y substitute for X, V
M(Y/X)
(, t, w) = 1
[V.] V
M
(., w) = 1 iff for every w e W, V
M
(, t, w) = 1
[V7
t
] V
M
(!t, w) = 1 iff for every t e T, V
M(t/t)
(, w) = 1
[Ve] V
M
(e, w) = 1 iff for some [
1
]
M
, [
n
]
M
, [t] and r
n
, such that [
1
]
M
, [
n
]
M
,
[t] e I (r
n
)
(w)
, [e]
M
= r
n
, [
1
]
M
, [
n
]
M
,

[t]
M.
[V7
e
] V
M
(!e, w) = 1 iff for every e e , V
M(e/e)
(, w) = 1
[V
e=
] V
M
(e = r
1

n
t) = 1 iff [e] = r
n
, [
1
]
M
, [
n
]
M
,

[t]
M
[Vee] V
M
(ee, w) = 1 iff for each e such that e ee, V
M
(e, w) = 1
[V7
ee
] V
M
(!ee, w) = 1 iff for every ee , , V
M(ee/ee)
(, w) = 1
[V] V
M
(e ee) iff [e]
M
e [ee]
M
,
[V] V
M
( , w) = 1 iff []
w
= []
w
, and V
M
(.( )) = 1
[V] V
M
( , w) = 1 iff for some , V
M
(( ), w) = 0, and {[]
w
^ []
w
} = []
w
[V] V
M
( , w) = 1 iff []
w
, []
w
,

and V
M
( , w) = 0
[V] V
M
( , w) = 1 iff for some , {[]
w
, []
w
} , []
w
and V
M
((( ^ ) ) = 0
There are derived clauses specifying the truth conditions for the non-primitive connectives as one would
expect.



229

Definitions of Validity
1. is valid in (W, D, I ) iff V
M
(, w) = 1 for every w e W and every assignment of .
2. is ML -valid (written
ML
) iff is valid in all ML models.
3. I ML semantically-implies (written I
ML
) iff for every ML model (W, D, I ), every w e
W, and for every variable assignment , V

(,, w) = 1 for each , e I, entails V

M
(, w) = 1.
















230

REFERENCES

Adams, R. (1981) Actualism and Thisness, Synthese 49: 3-41.
(1997) Things in Themselves, Philosophy and Phenomenological Research 57: 801- 825.
Anscombe, G.E.M. and Krner, S. (1964) Symposium: Substance, Aristotelian Society
(Supplementary) 38:69-90.
Antonelli, G.A. (2000) Proto-Semantics for Positive Free Logic, Journal of Philosophical Logic
29: 277-294.
Armstrong, D. M. (1968) A Materialist Theory of the Mind. London: Routledge & K. Paul.
(1978) Universals and Scientific Realism (Two Volumes) Cambridge: Cambridge
University Press.
(1983) What is a Law of Nature? Cambridge: Cambridge University Press.
(1989) A Combinatorial Theory of Possibility. Cambridge: Cambridge University Press.
(1997) A World of States of Affairs. Cambridge: Cambridge University Press.
Aspect, A., Dalibard, J., and Roger, G. (1982) Experimental Test of Bell's Inequalities Using
Time-Varying Analyzers, Physical Review Letters 49: 1804-1807.
Audi, P. (2012) A Clarification and Defense of the Notion of Grounding, in F. Correia and B.
Schnieder (eds.) Metaphysical Grounding: Understanding the Structure of Reality.
Cambridge: Cambridge University Press: 101-210.
Baker, R. (1967) Particulars: Bare, Naked, and Nude, Nos 1: 211-212.
Balashov, Y. (2002) What is a Law of Nature? The Broken Symmetry Story. Southern Journal of
Philosophy 40: 459-473.
Barcan, R. C. (1946) A Functional Calculus of First Order Based on Strict Implication, Journal
of Symbolic Logic 11:1-16.
(1947) The Identity of Individuals in a Strict Functional Calculus of Second Order, Journal
of Symbolic Logic 12:12-15.
Bauer, W. (2011) An Argument for the Extrinsic Grounding of Mass, Erkenntnis 74: 81-99
Bealer, G. (1982) Quality and Concept. Oxford University Press.
(1983) Completeness in the Theory of Properties, Relations, and Propositions, Journal of
Symbolic Logic 48: 415-426.
Bell, J.S. (1964) On the Einstein-Podolsky-Rosen paradox, Physics 1: 195200.
Bennett, K, (2006) Proxy Actualism, Philosophical Studies 129: 263-294.
(2011a) Construction Area (No Hard Hat Required), Philosophical Studies 154:79-104.
(2011b) By Our Bootstraps, Philosophical Perspectives 25:27-41.
Belnap, N. (1992) Branching Space-Time Synthese 92: 385-434.
Bird, A. (1998) Dispositions and Antidotes, Philosophical Quarterly 48: 227-234.
(2007) Nature's Metaphysics: Laws and Properties. Oxford: Oxford University Press.
Black, R. (2000) Against Quidditism, Australasian Journal of Philosophy 78:87104.
Borghini, A. and Williams, N. E. (2008) A Dispositional Theory of Possibility, Dialectica 62: 2141.
R. W. Byrne, R. W. and Whiten, A. (1988). Machiavellian Intelligence: Social Expertise and the
Evolution of Intellect in Monkeys, Apes, and Humans. Oxford University Press.
Butterfield, J. (1992) Review of B.C. van Fraassen, Quantum Mechanics: An Empiricists View,
Studies in the History and Philosophy of Modern Physics 24: 443-276.
Campbell, K. (1990) Abstract Particulars. Oxford: B. Blackwell.
Carnap, R. (1971) A Basic System of Inductive Logic, Part I, in Studies in Inductive Logic and
Probability, vol. I (Eds.), R. Carnap and R. C. Jeffery, University of California Press.
Cameron, R. P. (2005) Truthmaker Necessitarianism and Maximalism. Logique Et Analyse
48:43-56.

231

(2007) Lewisian Realism: Methodology, Epistemology, and Circularity Synthese
156:143 - 159.
(2008a) Truthmakers, Realism and Ontology, Royal Institute of Philosophy Supplements
83:107-128.
(2008b) How to Be a Truthmaker Maximalist, Nos 42:410 - 421.
(2010) The Grounds of Necessity, Philosophy Compass 5:348-358.
Chalmers, D. (1996) The Conscious Mind. Oxford University Press.
(2010) The Character of Consciousness. Oxford University Press.
Cheyne, C. and Pigden, C. (2006) Negative Truths From Positive Facts, Australasian Journal of
Philosophy 84:249 265.
Chisholm, R. M. (1967) Identity through Possible Worlds: Some Questions, Nos 1: 1-8.
Ciprotti, N. (2006) Minimising Existence. Or How to Stop Worrying and Love the Barcan
Formulae, Annali del Diparimento di Filosofia 12: 215-238.
Colbeck, R. Renner, R. (2011) No extension of quantum theory can have improved predictive power,
Nature Communications 411: arXiv:1005.5173v3
Contessa, G. (2010) Modal Truthmakers and Two Varieties of Actualism, Syntheses 174: 341- 353.
Correia, F. (2006) Generic Essence, Objectual Essence, and Modality, Nos 40:753767.
(2010) Grounding and Truth-Functions, Logique et Analyse 53: 251-79.
Craig, W. L., and Smith, Q. (Eds) (2007) Einstein, Relativity and Absolute Simultaneity. New York:
Routledge.
Darby, G. (2012) Relational Holism and Humean Supervenience, British Journal for the
Philosophy of Science 63: 773-788.
Davidson, D. (1967) The Logical Form of Action Sentences, in N. Rescher (ed.), The Logic of
Decision and Action, Pittsburgh: University of Pittsburgh Press: 8195.
(1969) The Individuation of Events, in N. Rescher (ed.), Essays in Honor of Carl G.
Hempel, Dordrecht: Reidel: 21634.
deRosset, L. (2010) Getting Priority Straight Philosophical Studies 149: 73-97.
(forthcoming) Grounding Explanations Philosophers Imprint.
(In Progress) On Weak Ground.
Deutsch, D. (1985) Quantum theory as a universal physical theory, International Journal of
Theoretical Physics 24: 1-41.
Deutsch, H. (1990a) Contingency and Modal Logic, Philosophical Studies 60:89-102.
(1990b) Real Possibility, Nos 24: 751-755.
Divers, J. (2002) Possible Worlds. Routledge.
Ellis, B. D. (1990) Truth and Objectivity. Oxford: Blackwell.
(2001) Scientific Essentialism. Cambridge University Press.
Esfeld, M. (2004) Quantum Entanglement and a Metaphysics of Relations, Studies in History and
Philosophy of Science Part B 35:601-617.
Everett, H. (1956/1973) The Theory of the Universal Wave Function, in B. De Witt and N. Graham
(eds.), The Many-Worlds Interpretation of Quantum Mechanics, Princeton NJ: Princeton
University.
(1957) Relative State Formulation of quantum mechanics, Review of Modern Physics 29:
454-462.
Fara, M. (2005) Dispositions and Habituals, Nos 39: 4382.
Fara, M. and Williamson, T. (2005) Counterparts and Actuality Mind 114:1-30.
Fodor, J. A. (1998) Concepts: Where Cognitive Science Went Wrong. Oxford University Press.
Fine, K. (1970) Propositional Quantifiers in Modal Logic, Theoria 36:336-346.
(1978) Model Theory for Modal Logic Part Ithe de Re/de Dicto Distinction. Journal of
Philosophical Logic 7:125-156.
(1982) First-Order Modal Theories III Facts, Synthese 53:43-122.

232

(1991) The Study of Ontology, Nos 25: 263-294.
(1994) Essence and Modality, Philosophical Perspectives 8: 1-16.
(2001) The Question of Realism, Philosophers' Imprint 1: 1-30.
(2012a) The Pure Logic of Ground, The Review of Symbolic Logic: 5: 1-25.
(2012b) A Guide to Ground, in F. Correia and B. Schnieder (eds.) Metaphysical
Grounding: Understanding the Structure of Reality. Cambridge: Cambridge University Press:
37-80.
Francescotti, R. (1999) How to Define Intrinsic Properties, Nos 33: 590-609.
(2007) Emergence, Erkenntnis 67:47 - 63.
Garson, J. (2001) Quantification in Modal Logic, in Gabbay, D., and Guenthner, F. (eds.)
Handbook of Philosophical Logic, 2ed edition, vol. 3, Dordrecht: D. Reidel: 267-323.
Goodman, N. (1955) Fact, Fiction, and Forecast, Cambridge, MA: Harvard University Press.
Hansson, T. (2007) The Problem(s) of Change Revisited, Dialectica 61: 265274.
Harris, R. (2010) Do Material Things Have Intrinsic Properties?, Metaphysica 11: 105-117.
Hauska, J. (2009) Dispositions Unmasked, Theoria 75: 304-335
Hawthorne, J. (2001) Intrinsic Properties and Natural Relations, Philosophy and
Phenomenological Research 63: 399-403
Hayaki, R. (2006) Contingent Objects and the Barcan Formula, Erkenntnis 64:75 - 83
Heil, J. (2000) Truth Making and Entailment, Logique and Analyse 43:231-242.
(2003) From an Ontological Point of View. Oxford University Press.
(2005) Dispositions, Synthese 144:343-356.
(2006) The Legacy of Linguisticism, Australasian Journal of Philosophy 84:233 244.
(2012) The Universe as we Find it. Oxford: Oxford University Press.
Heisenberg (1958) Philosophy and Physics. New York: Harper.
Hitchcock, C. and Knobe, J. (2009) Cause and Norm, Journal of Philosophy 106: 587-612.
Hughes and Cresswell (1996) A New Introduction to Modal Logic. New York: Routledge.
Hume, D. (1739/2000) A Treatise of Human Nature. D.F. Norton and M.J. Norton (eds.), Oxford
University Press.
Jackson, F. (1998) From Metaphysics to Ethics. Oxford University Press.
Jacobs, J. D., (2010) A Powers Theory of Modalityor, How I Learned to Stop Worrying and Reject
Possible Worlds Philosophical Studies 151:227-248.
(2011) Powerful Qualities, Not Pure Powers, The Monist 94: 81-102.
Jager, T (1982) An Actualistic Semantics for Quantified Modal Logic, Notre Dame Journal of Formal
Logic 23:335-349.
Johnson, W. E. (1921) Logic, Part I. Cambridge: Cambridge University Press.
Karakostas, V. (2008) Humean Supervenience in the Light of Contemporary Science, Metaphysica
10: 1-26.
Kim, J. (1973) Causation, Nomic Subsumption, and the Concept of Event, Journal of Philosophy 70:
217-236.
(1974) Noncausal Connections, Nos 8: 41-52
(1976) Events as Property Exemplifications, In M. Brand & D. Walton (eds.), Action Theory.
D. Reidel.
Kriegel, U. (2008) The Dispensability of (Merely) Intentional Objects, Philosophical Studies 141:
79-95.
Kripke, S. A. (1963) Semantical Considerations on Modal Logic, Acta Philosophica Fennica 16:83-
94.
(1972/1980) Naming and Necessity. Harvard University Press.
Ladyman, J. and Ross, D. (2007) Everything Must Go. Oxford: Oxford University Press.
Langton, R. (1998) Kantian Humility: Our Ignorance of Things in Themselves. Oxford University
Press.

233

(2004) Elusive Knowledge of Things in Themselves, Australasian Journal of Philosophy 82:
129136.
Langton, R. and Rochichaud, C. (2010) Ghosts in the World Machine? Humility and its Alternatives,
in A. Hazlett (ed.) New Waves in Metaphysics: 156-178. Palgrave Macmillan.
Leibniz, G. W. (1989) Philosophical Essays. Translated by R. Ariew and D. Garber. Indianapolis IN:
Hackett.
Leonard, H. S. and Goodman, N., (1940) The Calculus of Individuals and Its Uses, Journal of
Symbolic Logic 5: 45-55.
Leniewski, S., (1916) Podstawy oglnej teoryi mnogosci I, Moskow: Prace Polskiego Kola
Naukowego w Moskwie, Sekcya matematyczno-przyrodnicza (Eng. trans. by D. I. Barnett:
Foundations of the General Theory of Sets. I, in S. Leniewski, Collected Works, ed. S. J.
Surma, J. Srzednicki, D. I. Barnett, and F. V. Rickey, Dordrecht: Kluwer, 1992, Vol. 1: 129-173).
Lewis. D. K. (1968) Counterpart Theory and Quantified Modal Logic, The Journal of Philosophy 65:
11326.
(1970) How to Define Theoretical Terms, Journal of Philosophy 67:427-446.
(1973) Counterfactuals. Malden MA: Wiley-Blackwell.
(1983) New Work for a Theory of Universals, Australasian Journal of Philosophy 61: 343-
377.
(1986a) On the Plurality of Worlds. Malden MA: Wiley-Blackwell.
(1986b). Philosophical Papers Vol. II Oxford University Press.
(1991) Parts of Classes. Malden MA: Wiley-Blackwell.
(1992) Critical Notice of David Armstrong, A Combinatorial Theory of Possibility,
Australasian Journal of Philosophy 70:211-224.
(1993) Mathematics in Megethology, Philosophia Mathematica 1:3-23.
(1994) Humean Supervenience Debugged, Mind 103: 473-490.
(1995) Should a Materialist Believe in Qualia? Australasian Journal of Philosophy 73:140-
44.
(1997) Finkish Dispositions, Philosophical Quarterly 47 (187):143-158.
(1999) Papers in Metaphysics and Epistemology. Cambridge: Cambridge University Press.
(2002) Tensing the Copula, Mind 111:1-14.
(2004) How Many Lives Has Schrodinger's Cat?, Australasian Journal of Philosophy 82:3-
22.
(2009) Ramseyan Humility. In D. Braddon-Mitchell and R. Nola (eds.), Conceptual
Analysis and Philosophical Naturalism. Mit Press.
Linde, A. (1985) The Universe: Inflation out of Chaos. New Scientist 105: 14-18.
Linsky, B, and Zalta, E.N. (1994) In Defense of the Simplest Quantified Modal Logic, Philosophical
Perspectives 8:431-458.
(1996) In Defense of the Contingently Nonconcrete, Philosophical Studies 84:283-294.
Livengood, J. and Machery, E. (2007) The Folk Probably Don't Think What You Think They Think:
Experiments on Causation by Absence, Midwest Studies in Philosophy 31: 107127.
Mackie, P. (2006) How Things Might Have Been: Individuals, Kinds, and Essential Properties. Oxford:
Oxford University Press.
Marcus, R. B. (1961) Modalities and Intensional Languages, Synthese 13:303-322.
(1967) Essentialism in Modal Logic, Nos 1:91-96.
(1986) Possibilia and Possible Worlds, Grazer Philosophische Studien, 25-26:107-133.
Martin, C. B. (1994) Dispositions and Conditionals, Philosophical Quarterly 44:1-8.
(1996) How It Is: Entities, Absences and Voids Australasian Journal of Philosophy 74:57
65.
(1997) On the Need for Properties: The Road to Pythagoreanism and Back, Synthese
112:193-231.
Martin, C.B., and Heil, J. (1999) The Ontological Turn Midwest Studies in Philosophy 23:34 60.

234

Morganti, M. (2009) A New Look at Relational Holism in Quantum Mechanics, Philosophy of Science
76: 1027-1038.
Maudlin, T. (2007) The Metaphysics within Physics. Oxford University Press.
McKitrick, J. (2003) The Bare Metaphysical Possibility of Bare Dispositions, Philosophy and
Phenomenological Research 66: 349369.
McMichael, A. (1983) A Problem for Actualism About Possible Worlds, Philosophical Review 92:49-
66.
Menzel, C. (1987) Theism, Platonism, and the Metaphysics of Mathematics, Faith and Philosophy 4:
365-382.
(1991) Temporal Actualism and Singular Foreknowledge, Philosophical Perspectives 5:475-
507.
(1993) The Proper Treatment of Predication in Fine-Grained Intensional Logic,
Philosophical Perspectives 7: 61-87.
(2008) Actualism, Stanford Encyclopedia of Philosophy.
Merricks, T. (2001) Objects and Persons. Oxford University Press.
Molnar, G. (2003) Powers: A Study in Metaphysics. Oxford University Press.
Moreland, J. P. (1998) Theories of Individuation: A Reconsideration of Bare Particulars, Pacific
Philosophical Quarterly 79: 251263.
Movahed (2006) http://www.mullasadra.org/new_site/english/Paper%20Bank/Logic/
movahed.htm.
Mller, T. (2005) Probability Theory and Causation: A Branching Space-Times Analysis, British
Journal for the Philosophy of Science 56: 487-520.
(2011) Probabilities in Branching Structures, In D. Dieks. et el (eds.), Explanation,
Prediction, and Confirmation: 109-21. Utrecht: Springer Netherlands.
(2012) Branching in the Landscape of Possibilities, Synthese 188:41-65.
Mumford, S. (2006) The Ungrounded Argument, Synthese 149: 471 489.
Mumford, S. and Anjum, R. L. (2011) Getting Causes from Powers. Oxford University Press.
Mulligan, K., Simons, P., and Smith, B. (1984) Truth Makers, Philosophy and Phenomenological
Research 44:287-321
(2006) What's Wrong with Contemporary Philosophy?, Topoi 25: 63-67.
Mundy, B. (1989) Elementary Categorial Logic, Predicates of Variable Degree, and Theory of
Quantity, Journal of Philosophical Logic 18:115 - 140.
Nelson, M. (2007) Ways an Actualist Might Be, Philosophical Studies 133:455-471.
(2009) The Contingency of Existence, in Samuel Newlands & Larry M. Jorgensen (eds.),
Metaphysics and the Good: Themes from the Philosophy of Robert Merrihew Adams. Oxford:
Oxford University Press.
Nelson, M. and Zalta, E.N. (2009) Bennett and Proxy Actualism, Philosophical Studies
142:277-292.
Ney, A. (2008) Physicalism as an Attitude, Philosophical Studies 138:1 - 15.
(2010) Are There Fundamental Intrinsic Properties?, in A. Hazlett (ed.) New Waves in
Metaphysics: 219-239. Palgrave Macmillan.
(2012) Neo-Positivist Metaphysics, Philosophical Studies 160:53-78.
Nolan, D. (2009) Platitudes and Metaphysics, in D. Braddon-Mitchell & R. Nola (eds.),
Conceptual Analysis and Philosophical Naturalism. Mit Press.
Nolt, J. (2010) Free Logic, Stanford Encyclopedia of Philosophy.
O'Connor, T. and Jacobs, J (2003) Emergent Individuals Philosophical Quarterly 53: 540- 555.
Parsons, T. (1969) Essentialism and Quantified Modal Logic, Philosophical Review 78:35-52
(1995) Ruth Barcan Marcus and the Barcan Formula, in Sinnot-Armstrong, Raffman, and
Asher (ed.), Modality, Morality, and Belief: Essays in Honor of Ruth Barcan Marcus. Cambridge:
Cambridge University Press.
Paul, L.A. (2006) In Defense of Essentialism, Philosophical Persepectives 20: 333-372.

235

(2012) Building the World from its Fundamental Constituants, Philosophical Studies 158:
221-256
Peramatzis, M. (2011) Priority in Aristotles Metaphysics. Oxford University Press.
Plantinga, A. (1974) The Nature of Necessity. Oxford: Oxford University Press.
(1976) Actualism and Possible Worlds, Theoria 42:139-160.
(1983) On Existentialism, Philosophical Studies 44:1 - 20.
(1986) Two Concepts of Modality: Modal Realism and Modal Reductionism, Journal of
Philosophy 83: 189-231.
Plato (1961) The Collected Dialogues E. Hamilton and H. Cairns (eds.), Princeton University Press.
Prior, A. N. (1956) Modality and Quantification in S, Journal of Symbolic Logic 21:60-62.
(1957) Time and Modality. Oxford: Oxford University Press.
(1968) Past, Present, and Future. Oxford: Oxford University Press.
Pruss. A. (2002) The Actual and the Possible. In Richard M. Gale (ed.), Blackwell Guide to
Metaphysics, Oxford: Blackwell.
Quine, W.V.O. (1948) On What There Is, Review of Metaphysics 2:21-38.
(1950) Identity, Ostension, and Hypostasis, Journal of Philosophy 47: 621-633.
(1953/1961) Reference and Modality. In W. O. V. Quine, From a Logical Point of View.
Harvard University Press.
(1969) Ontological Relativity and Other Essays. Columbia University Press.
Rawls, J. (1971/1999) A Theory of Justice, Cambridge, MA: Harvard University Press.
(1974), The Independence of Moral Theory, Proceedings and Addresses of the American
Philosophical Association 47: 52.
Resnik, M. D. (1990) Between Mathematics and Physics, PSA: Proceedings of the Biennial
Meeting of the Philosophy of Science Association 1990: 369-378.
Rosen, G. (2010) Metaphysical Dependence: Grounding and Reduction, in B. Hale and A.
Hoffmann (eds.), in Modality: Metaphysics, Logic, and Epistemology. Oxford University Press.
(2012) "Abstract Objects", The Stanford Encyclopedia of Philosophy, E. N. Zalta (ed.), URL =
<http://plato.stanford.edu/archives/spr2012/entries/abstract-objects/>.
Ross, D. (2005) Economic Theory and Cognitive Science: Microexplanation. Cambridge MA: MIT
Press.
Rovelli, C. (1996) Relational Quantum Mechanics, International Journal of Theoretical Physics
35: 1637-1678.
(1998) Loop Quantum Gravity, Living Reviews in Relativity 1.
Rumfitt, I. and Williamson, T. (1999) Logic and Existence, Aristotelian Society
Supplementary Volume 73:181203.
Ryle, G. (1949) The Concept of Mind. Hutchinson and Co.
Samuels, R. (2012) Science and Human Nature, Royal Institute of Philosophy Supplement 70: 1-28.
Schaffer, J. (2005) Quiddistic Knowledge, Philosophical Studies 123: 1-32.
(2008) Truthmaker Commitments, Philosophical Studies 141: 7-19.
(2009) On What Grounds What, In David Manley, David J. Chalmers & Ryan Wasserman
(eds.), Metametaphysics: New Essays on the Foundations of Ontology. Oxford University Press.
(2010) Monism: The Priority of the Whole, Philosophical Review 119: 31-76.
Schnieder, B. (2011) A Logic for Because, The Review of Symbolic Logic 4: 445-465.
Searle, J. (1959) On Determinables and Resemblance II, Proceedings of the Aristotelian Society
(Supplementary Volume) 33: 141158.
(1980) Minds, Brains and Programs, Behavioral and Brain Sciences 3: 417-57.
Sider, T. (2001) Four Dimensionalism: An Ontology of Persistence and Time. Oxford:Oxford
University Press.
(2002) The Ersatz Pluriverse. Journal of Philosophy 99:279-315.
(2006) Bare Particulars. Philosophical Perspectives 20 (1):387397.
(2009) Williamson's Many Necessary Existents. Analysis 69 (2):250-258.

236

(2010) Logic for Philosophy. Oxford University Press.
(2011). Writing the Book of the World. Oxford University Press.
Shoemaker, S. (1980) Causality and Properties, In Peter van Inwagen (ed.), Time and Cause. D.
Reidel.
(1998) Causal and Metaphysical Necessity, Pacific Philosophical Quarterly 79: 5977.
Skyrms, B. (1981) Tractarian Nominalism, Philosophical Studies 40: 199-206.
Simchen, O. (forthcoming) The Barcan Formula in Metaphysics, Theoria.
Sorkin, R.D. (1990) First Steps with Causal Sets, in R. Cianci, R. de Ritis, M. Francaviglia, G. Marmo,
C. Rubano, P. Scudellaro (eds.), General Relativity and Gravitational Physics: 6890.
Stalnaker, R.C. (1976) Possible Worlds, Nos 10: 65-75.
(1979) Anti-Essentialism, Midwest Studies in Philosophy 4:343-355.
(2003) Ways a World Might Be: Metaphysical and Anti-Metaphysical Essays. Oxford
University Press.
(2011) Mere Possibilities. Princeton University Press.
Strawson, G. (2006) Realistic Monism - Why Physicalism Entails Panpsychism, Journal of
Consciousness Studies 13: 3-31.
(2008) The Identity of the Categorical and the Dispositional, Analysis 68: 271-282.
Surez, M. (2011) Four Theses on Probabilities, Causes, Propensities, In M. Surez (Ed.),
Probabilities, Causes and Propensities in Physics: 1-40. New York: Springer.
Sullivan, M. (2012a) Problems for Temporary Existence in Tense Logic, Philosophy Compass 7:43-
57.
(2012b) The Minimal A-Theory, Philosophical Studies 158:149-174.
Swinburne, R. (2007) Cosmic Simultaneity, in W. L. Craig and Q. Smith.
Tarski, A. (1944) The Semantic Conception of Truth: And the Foundations of Semantics,
Philosophy and Phenomenological Research 4:341-376.
Tegmark, M. (1998) The Interpretation of Quantum Mechanics: Many Worlds or Many Words?
Fortschr. Phys., 46, 855-62.
(2003) Parallel Universes. In John. D. Barrow, P.C. Davies, and C.L. Harper Jr. (Eds.) Science
and Ultimate Reality. Cambridge: Cambridge University Press: 459-91.
Teller, P. (1986) Relational Holism and Quantum Mechanics, British Journal for the Philosophy of
Science 37: 71-81.
(1990) Prolegomenon to a Proper Interpretation of Quantum Field Theory, Philosophy of
Science 57: 594-618.
Thomason, R.H. (1984) Combinations of Tense and Modality. In D. Gabby and F. Guenthner
(Eds.) The Handbook of Philosophical Logic (1st Ed.) Vol. 2. Boston MA: Syntheses
Library:135-66.
van Fraassen, B. C. (1969) Facts and Tautological Entailments, Journal of Philosophy 66: 477- 487.
(1989) The Scientific Image. Oxford University Press.
van Inwagen, P. (2011) Relational vs. Constituent Ontologies, Philosophical Perspectives 25: 389-
405.
Weinberg, J. M., Nicholas, S. and Stich, S. (2001) Normativity and Epistemic Intuitions,
Philosophical Topics, 29: 429-460.
Wiggins, D. (1967) Identity and Spatio-Temporal Continuity. Oxford: Blackwell.
(1980/2001) Sameness and Substance. Harvard University Press.
Wigner (1961) Remarks on the Mind-Body Problem, in The Scientist Speculates, I.G. Good (ed.),
London: W. Heinemann.
Williamson, T. (1998) Bare Possibilia, Erkenntnis 48:257-273.
(1999) Truthmakers and the Converse Barcan Formula, Dialectica 53:253270.
(2000) The Necessary Framework of Objects, Topoi 19:201-208.
(2002) Necessary Existents, In A. O'Hear (ed.), Logic, Thought, and Language.
Cambridge: Cambridge University Press.

237

(2003) Everything, Philosophical Perspectives 17:415465.
(2006) Stalnaker on the Interaction of Modality with Quantification and Identity, In J. J.
Thomson and A. Byrne (eds.), Content and Modality: Themes from the Philosophy of Robert
Stalnaker. Oxford University Press.
(2007) How Probable is an Infinite Sequence of Heads?, Analysis 67 (295):173-180.
(2010) Necessitism, Contingentism, and Plural Quantification, Mind 119: 657-748.
(forthcoming) Barcan Formulas in Second-Order Modal Logic, In Themes from Barcan
Marcus. Ontos Verlag.
(2013) Modal Logic as Metaphysics. Oxford: Oxford University Press.
Wilson, J.M. (forthcoming) Humes Dictum and Metaphysical Modality: Lewis
Combinatorialism, In Barry Loewer & Jonathan Schaffer (eds.), The Blackwell
Companion to David Lewis. Malden, MA: Blackwell.
Wittgenstein, J. (1921/1961) Tractatus Logico-Philosophicus. D.F. Pears and B.F. McGuinness. New
York: Routledge
Yablo, S. (1982) Grounding, Dependence, and Paradox Journal of Philosophical Logic 11: 117-
137.
(1993) Is Conceivability a Guide to Possibility? Philosophy and Phenomenological
Research 53:1-42.
Zalta, E.N. (1983) Abstract Objects: An Introduction to Axiomatic Metaphysics, D. Reidel.
(1988) Intensional Logic and the Metaphysics of Intentionality. The MIT Press.
(1999) Principia Metaphysica. http://mally.stanford.edu/publications.html.
(2006) Essence and Modality. Mind 115 (459):659-693.





















238

VITA

Adam James Labecki


EDUCATION
Dalhousie University
M.A. in Classics, October 2006
Dalhousie University
B.A. Combined Honors in Philosopy and Contemporary Studies

EMPLOYMENT
University of Kencucky
Instructor: Spring 2009 Present; Teaching Assistant Fall 2007 Fall 2008
Dalhousie University
Teaching Assistant Spring 2005- Spring 2006
University of Kings College
Teaching Assistant Spring 2004 Spring 2006

AWARDS
University of Kencucky Graduate Shool Fellowship, Spring 2012

PUBLICATIONS
The One and the Many: Part II: The Many. Dionysius 25: 129-152 (2007)
The One and the Many: Part I: The One. Dionysius 24: 75-98 (2006)