Full Chapter A Metaphysics of Platonic Universals and Their Instantiations Shadow of Universals Jose Tomas Alvarado PDF
Full Chapter A Metaphysics of Platonic Universals and Their Instantiations Shadow of Universals Jose Tomas Alvarado PDF
Full Chapter A Metaphysics of Platonic Universals and Their Instantiations Shadow of Universals Jose Tomas Alvarado PDF
https://textbookfull.com/product/human-universals-donald-e-brown/
https://textbookfull.com/product/ethnocentric-political-theory-
the-pursuit-of-flawed-universals-bhikhu-parekh/
https://textbookfull.com/product/the-problem-of-universals-in-
early-modern-philosophy-1st-edition-stefano-di-bella/
https://textbookfull.com/product/yoga-meditation-and-mysticism-
contemplative-universals-and-meditative-landmarks-kenneth-rose/
Expressing the Self: Cultural Diversity and Cognitive
Universals Minyao Huang (Editor)
https://textbookfull.com/product/expressing-the-self-cultural-
diversity-and-cognitive-universals-minyao-huang-editor/
https://textbookfull.com/product/statistical-universals-of-
language-mathematical-chance-vs-human-choice-1st-edition-kumiko-
tanaka-ishii/
https://textbookfull.com/product/on-the-genealogy-of-universals-
the-metaphysical-origins-of-analytic-philosophy-first-edition-
edition-fraser-macbride/
https://textbookfull.com/product/knowledge-and-ignorance-of-self-
in-platonic-philosophy-james-m-ambury/
https://textbookfull.com/product/metaphysics-of-science-a-
systematic-and-historical-introduction-1st-edition-markus-
schrenk/
Synthese Library 428
Studies in Epistemology, Logic, Methodology,
and Philosophy of Science
A Metaphysics
of Platonic
Universals
and their
Instantiations
Shadow of Universals
Synthese Library
Editor-in-Chief
Otávio Bueno, Department of Philosophy, University of Miami,
Coral Gables, USA
A Metaphysics of Platonic
Universals and their
Instantiations
Shadow of Universals
José Tomás Alvarado
Instituto de Filosofía
Pontificia Universidad Católica de Chile
Macul – Santiago, Chile
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland
AG 2020
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by
similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Acknowledgments
This work was written during the execution of the research project Fondecyt
1160001 (2016–2019; Conicyt, Chile). It is, however, the fruit of more than
10 years’ work. The ideas presented here have been developed with the support of
the Fondecyt research projects 1070339 (2007–2008), 1090002 (2009–2011), and
1120015 (2012–2014). Many more people deserve my gratitude than I am able to
include here. In particular, I must mention the Colloquia of Analytic Metaphysics
that, since its first version in the Argentine Society of Philosophical Analysis
(SADAF) in Buenos Aires in 2008, has continued to meet every 2 years. Many of
the theses that appear in this book found their first formulations in those colloquia. I
must thank Ezequiel Zerbudis, Gonzalo Rodriguez-Pereyra, Juan Larreta (requiescat
in pace) Guido Imaguire, Sebastián Briceño, Carlo Rossi, Robert Garcia, and
Horacio Banega for their comments, criticisms, and suggestions on those memorable
occasions. In October of 2017, Sebastián Briceño organized a seminar in the
Department of Philosophy of the University of Concepción that was dedicated to
the discussion of the first version of this book. I am very grateful to Sebastián
Briceño and Javier Vidal for their sharp observations that greatly helped to improve
this draft. Marcelo Boeri, Juan Manuel Garrido, and Matthew Tugby have kindly
read the manuscript and made me see several errors that I have been able to rectify.
An anonymous referee for Springer made many useful observations that have
improved the book in many parts. Finally, I thank my colleagues from the Institute
of Philosophy of the Pontificia Universidad Católica de Valparaíso and from the
Institute of Philosophy of the Pontificia Universidad Católica de Chile who have
created a stimulating environment for philosophical work; and I thank the genera-
tions of students who have been exposed—in one way or another—to the ideas
explored here.
v
Contents
1 Introduction [§ 1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Properties, Universals, Tropes [§ 2] . . . . . . . . . . . . . . . . . . . . . 2
1.2 Possible Worlds, Grounding, Dependence [§ 3–4] . . . . . . . . . . . 8
1.3 Intrinsic and Extrinsic Properties, Mereology, Concrete
and Abstract [§ 5–7] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 Summary of What Is to Come and some Nomenclature [§ 8] . . . 15
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Part I Universals
2 Theoretical Roles for Universals [§ 9] . . . . . . . . . . . . . . . . . . . . . . . 23
2.1 One Over Many [§ 10] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2 Many Over One [§ 11] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3 Objective Resemblances [§ 12] . . . . . . . . . . . . . . . . . . . . . . . . 27
2.4 Causality [§ 13] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.5 Natural Laws [§ 14] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.6 Inductive Practices [§ 15] . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3 The Superiority of Universals Over Resemblance
Nominalism [§ 16] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.1 Resemblance Nominalism [§ 17] . . . . . . . . . . . . . . . . . . . . . . . 48
3.2 Primitive Facts of Resemblance [§ 18] . . . . . . . . . . . . . . . . . . . 54
3.3 How Would We Have Epistemic Access to Such
Resemblance Facts? [§ 19] . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.4 Determinate and Determinable [§ 20] . . . . . . . . . . . . . . . . . . . . 61
3.5 Natural Laws and Inductions [§ 21] . . . . . . . . . . . . . . . . . . . . . 65
3.6 Causal Powers [§ 22] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.7 A Vicious Regress [§ 23] . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.8 Modal Consequences [§ 24] . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
vii
viii Contents
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Chapter 1
Introduction
Abstract This first chapter presents the main line of argumentation that is going to
be followed in the book for the defence of an ontology of Platonic universals and
their instantiations. Several important concepts and theses for that argumentation are
introduced, like ‘property’, ‘universal’, ‘trope’, ‘possible world’, ‘grounding’,
‘dependence’, ‘intrinsic/extrinsic property’, ‘mereological sum’ and ‘abstract/
concrete’.
§ 1. If we consider everything in the broadest possible sense, we may say that certain
entities are grounded on others (cf. § 4), while some entities are fundamental because
they are ungrounded. The central thesis of this work is that everything on the
fundamental level is either a transcendent universal or a trope. Many other categories
of entities can be admitted, such as sets, mereological fusions, states of affairs,
events, concrete structures, and—especially—particular objects; but they are onto-
logically derivative. This thesis has never enjoyed great acceptance, and a long and
relatively intricate journey will be undertaken to defend it here. The strategy to
justify this central thesis can be summarized in the following sub-theses:
(A) There are universals
(B) Every universal is independent of its instantiations
(C) There are particular objects
(D) Every particular object is a bundle of tropes
Sub-theses (A) and (B) can be taken relatively independently of sub-theses
(C) and (D). The defence of (A) and (B) will be made in Parts I and II of this
work (Chaps. 2, 3, 4, 5, 6, 7, 8, 9, and 10, §§ 9–77). The defence of (C) and (D) will
be made in Part III (Chaps. 11, 12, and 13, §§ 78–94). That is, the thesis to be
defended is that there are universal properties (sub-thesis (A)), but that—contrary
to what many of its defenders have assumed—these universal properties do not need
to be instantiated to exist (sub-thesis (B)). Much of this work will be dedicated to
justifying, then, that there are universals that are not instantiated. When it comes to
an understanding of the nature of particular objects, on the other hand, I will argue
that these should be understood as complexes of the instantiations of transcendent
© The Editor(s) (if applicable) and The Author(s), under exclusive license to 1
Springer Nature Switzerland AG 2020
J. T. Alvarado, A Metaphysics of Platonic Universals and their Instantiations,
Synthese Library 428, https://doi.org/10.1007/978-3-030-53393-9_1
2 1 Introduction
universals (sub-thesis (D); sub-thesis (C) does not require a justification). These
‘instantiations’ are tropes or particular properties. This ontology of tropes will be
explained in Part III. When considering the nature of particular objects it has been
usual to argue that there are two mutually exclusive alternatives: they are understood
either as being made up of a substratum that has properties, or as a cluster or bundle
of properties without a substratum. I will here advance the claim that there is no such
opposition and, moreover, will defend an ontology that understands particular
objects as tropes, but at the same time as having a substratum (see §89).
The central thesis to be defended in this work is that, at the fundamental level,
everything that exists is a universal or an instantiation of a universal. The ordinary
world of objects and events that happen to these objects is, in some sense, a world of
‘shadows’ of the universals of which the particular properties or tropes are essen-
tially an instantiation. Talk about ‘shadows of universals’ has Platonic resonances, of
course (cf. Plato (1982), VII, 514a-521b), but it is not used here except as a suitable
metaphor. It is not intended here to argue, for example, that the existence of concrete
objects and the particular properties that make them is an ‘illusion’, or that ‘really’
these objects do not exist, or that their existence is of a ‘minor degree’. Rather, the
metaphor is used here in the key sense that a shadow projects information about what
causes it. The particular objects are instantiations of universals. The reality of the
world of particular objects also ‘projects’ information from a different realm: a realm
of universals.
In this introductory chapter, along with a global presentation of the central
defence strategy to be employed, some notions of importance will be presented,
such as ‘property’, ‘universal’, ‘trope’, ‘possible world’, or ‘part’.
1
Although these are not very good examples, as will be explained below. There is probably no
unique real property of ‘being red’ or of ‘being yellow’. The closest thing to such a property would
be infinite pluralities of determined colour properties, i.e. of maximally specific colour tones. These
pluralities are vague, as there will be certain determined colour tones that will not clearly be red
tones or non-red tones. It is doubtful, then, that there are such pluralities.
1.1 Properties, Universals, Tropes 3
considers what it is that grounds such natures—if there is anything on which they are
grounded. In the following, a characteristic, feature, or determination of an object
that is supposed to be numerically different from the object that possesses it will be
called a ‘property’. The connection between a property and the object that owns it
will be designated as its “instantiation”, however such a ‘connection’ is finally
understood.2 An ontology that postulates properties, therefore, postulates that the
objects that are usually presented to us have a specific ontological structure—at least
in the sense in which the expression “property” is used here. What is presented to us
usually must be a ‘complex’ that must include properties that are numerically
different from the object that has them. Not all philosophers use this terminology
to designate a ‘property’. David Lewis (cf. 1983, 10) and Gonzalo Rodriguez-
Pereyra (cf. 2002, 15–17), for example, use “property” to designate whatever it is
that satisfies the theoretical functions usually attributed to universals. The theories
that they have proposed, then, are presented as theories about the nature of ‘prop-
erties’, even though according to the terminology employed here neither Lewis nor
Rodriguez-Pereyra postulates the existence of properties. For simplicity, it is
assumed that the properties in question will be monadic, but everything that is said
should be considered applicable mutatis mutandis for relations, that is, for dyadic
properties or properties with a larger number of arguments (or ‘adicity’). When the
distinction between monadic properties and relationships is relevant, the distinction
will be made explicitly.
A theory that rejects the existence of properties in the indicated sense is a form of
nominalism. Here, too, there is considerable terminological variability. In the phi-
losophy of mathematics, any theory that rejects the existence of abstract entities,
such as sets or classes (see, for example, Goodman and Quine 1947) has been called
“nominalist”. A traditional nominalist strategy to eliminate universals is to propose
classes of objects or natural classes of objects that can fulfil their functions. A class is
an abstract entity, however, of those that the ‘nominalists’ in the philosophy of
mathematics reject. Any theory that rejects the existence of universals has also been
called “nominalist” (see, for example, Armstrong 1978a, 138). According to the
terminology that is followed here, however, one can defend the existence of prop-
erties that are not universal.
Properties can be conceived as having a universal or a particular nature. A
universal property is a property that can be instantiated in a plurality of exemplifi-
cations at the same time. A universal can be shared by many objects. Unlike a
universal, a trope is a property of particular and not universal character. The
terminology here has also been very variable: they have been called “modes”,
“moments”, “abstract particulars”, “perfect particulars”, “concrete properties”, or
“instances of property”. From now on, they will be designated as “tropes”. A trope
cannot be shared by several objects. For a defender of universals, for example, the
mass of an electron is numerically the same property as the mass of another. For a
2
Later on (see § 79), other additional distinctions required for an adequate understanding of
particular objects will be introduced.
4 1 Introduction
defender of tropes, on the other hand, the mass of an electron is numerically different
from the mass of another electron, if these electrons are numerically different.3
Tropes can be perfectly similar to each other, but this is another matter.
There are two ways in which universals can be understood: as immanent or
‘Aristotelian’ universals, or as transcendent or ‘Platonic’ universals. The denomina-
tions “Aristotelian” or “Platonic” should not be taken too seriously here. One cannot
confidently ascribe to Aristotle the postulation of immanent universals.4 Neither can
one confidently attribute to Plato the postulation of transcendent universals (for a
general presentation, see Ross 1953), as those universals have been conceived in the
recent discussion—and indeed, the acceptance of a theory like the one that will be
defended here could not be attributed to Plato. An immanent universal is a universal
that exists only if it has instantiations. A transcendent universal is a universal that
does not require instantiations to exist. In other words, the difference between one
and another conception of universals has to do with the acceptance or rejection of
this principle (see Armstrong 1978a, 113):
[Principle of Instantiation] It is necessary for all n-adic universal U that
necessarily at least n objects instantiate U.5
The defenders of immanent universals accept this Principle of Instantiation, while
the defenders of transcendent universals reject it. Immanent universals are, therefore,
ordinarily contingent entities. Except for the case of essential properties for an
object, the fact that an object instantiates a property is a contingent fact. Still,
however, if it were an essential property for an object, if the object in question
were a contingent and not a necessary entity, then the property would also be
contingent and not necessary. In effect, as an immanent universal U exists only in
the possible worlds6 where it is instantiated, even if it is an essential property of an
object, in the worlds where that object does not exist, U will not either—assuming
that no other objects in those worlds instantiate U. For a defender of immanent
universals, the only cases of necessary universals would be the cases of essential
properties of necessary objects that invariably exist in all possible worlds. Necessary
universals could be got, perhaps, in the case of essential properties of mathematical
objects, if any, and the essential properties of God.
The situation is very different when it comes to transcendent universals. It would
not be inconsistent to suppose that these universals are contingent entities. All that is
3
An assumption that should always be taken with extreme caution when dealing with entities at the
quantum level. Cf. French and Krause 2006, 84–197.
4
Indeed, it has been the subject of discussion whether Aristotle admits universal or particular forms
in Metaphysics, especially about Z 13. Cf. for general presentations, Wedin 2000; Lewis 2013.
5
Armstrong’s original formulation (see 1978a, 113) does not include modal operators, but clearly,
he does not propose the principle as a happy contingent coincidence affecting all actually existing
universals. The principle obtains for the defender of an Aristotelian conception of universals due to
the essential nature of universals, i. e. as a matter of metaphysical necessity. The insertion of modal
operators in the principle above makes explicit this modal character.
6
On the concept of ‘possible world’, cf. § 3 below.
1.1 Properties, Universals, Tropes 5
7
There are other more exotic hypotheses discussed in the literature that will not be discussed here. It
has been argued, for example, that a particular object could also be multi-located entirely in
different regions of space because an object could travel in time to the past and occupy a different
region of space than it occupies at any given time (Ehring 2011, 27–30). It is not intended to argue
here that the differences between universals and individuals concerning their spatio-temporal
location are the way to analyse the difference between universal and particular.
6 1 Introduction
There has been a debate about how to analyse the distinction between universals
and particulars. Many criteria have been proposed, and many of them have been
unsatisfactory. One might think, for example, that the difference between a universal
and a particular consists in the fact that the universal—in the manner of a Fregean
function—is essentially incomplete and required to be saturated by an object.
Ramsey, however, showed that one could think analogously of objects as ‘incom-
plete’ and as ‘requiring saturation’ by one property or another (see Ramsey 1925).8
One might think, too, that the way to analyse the difference between universal and
particular is that individuals may enter into connections with an arbitrary number of
other objects and universals to form states of affairs. A universal, on the other hand,
seems to be able to enter into connection only with a fixed number of objects,
corresponding to its adicity, that instantiate it (cf. Armstrong 1997, 168; MacBride
2005, 571–572). It turns out, however, that one cannot exclude a priori the existence
of multigrade universals, that is, universals that can be instantiated in arbitrary
pluralities of objects. Suppose there were a universal of to surround—a hypothesis
not very plausible, but sufficient to clarify this point. It can be given the state of
affairs of being Peter, John, and James surrounding x, but also the state of affairs of
being Peter, John, James, and Andrew surrounding x, or the state of affairs of being
Peter and John surrounding x, etc. The same universal seems to be integrating states
of affairs with two, three, four, or n different objects. The fact that there is no way to
analyse the difference between universals and particulars, however, should not be
taken as a reason to argue that there is no difference between them. Not everything
can be analysed. The difference between universal and particular might be of such a
fundamental nature that there are no more basic conceptual resources from which it
could be made clear. It would not be strange if such a thing should happen. In the
following it is assumed that the distinction is simply primitive.9
A distinction has frequently been postulated between sparse and abundant prop-
erties, at least since it was presented in these terms by David Lewis (cf. 1983,
11–14). A traditional way of postulating universal properties has been to argue
8
Suppose that the proposition that Socrates is wise corresponds to a state of affairs. The semantic
components of the proposition correspond to the ontological components of the state of affairs. In
the proposition, we can distinguish a function x is wise in which the ‘x’ indicates the empty place of
the variable that requires saturation or to be bounded by a quantifier to generate a complete
proposition. Socrates seems to be ‘complete’. This contrast, however, seems excessively deter-
mined by the grammatical difference between predicates and names. Ramsey (1925) argued that
one could just as well represent Socrates as Socrates is F, where F is a variable that needs to be
saturated to form a complete proposition. If the ‘unsaturated’ character implies a certain ontological
precariousness and is not merely semantic, the same seems to happen with universals. Universals
cannot constitute states of affairs independently of objects—at least in the immanentist conception
of universals—nor can objects form states of affairs independently of universals.
9
Which does not rule out, of course, that some appropriate way of doing the analysis should appear.
For example, it has been proposed that the difference between universal and particular is given by
the fact that it is metaphysically impossible for universals to be perfectly similar to each other or
indiscernible among themselves (cf. Ehring 2011, 30–40). On the other hand, it is possible that there
are particulars which are indiscernible between themselves, whether they are objects or tropes.
1.1 Properties, Universals, Tropes 7
that they are the semantic value of the predicates of our language or that they are the
content of our concepts. Each possible predicate will be correlated with a ‘property’,
as well as each concept of a possible thinking subject. For a natural language like
English, which allows us to build denumerably infinite many10 different predicates
by the recursive application of compositional rules of grammatical construction,
there are infinite predicates that seem to select extremely heterogeneous things. For
example, there is a predicate like ‘being a cat and being examined before the year
3,000, or being a galaxy and being examined after the year 3,000’. In this conception
there is exactly one class of all the things that instantiate the ‘property’ which is the
meaning of the predicate in question. This class includes cats and galaxies. Examples
like this can be multiplied ad nauseam. There are certain theoretical functions that
properties specified in this way have been expected to satisfy. The discussion in this
work, however, will concentrate on the so-called ‘sparse properties’. These are
‘sparse’ because in comparison to the ‘abundant’ ones they form a minority. These
properties are those that are strictly necessary to ground the objective resemblances
and the causal powers of objects. The way in which we can determine if there is a
property in this sense is not through reflection on the contents of our thought and our
language, but rather through reflection on what are the objective resemblances
between the objects and what causal powers they have—according to our best
evidence regarding such objects, which will ordinarily be empirical. It is not the
task of the philosopher to determine, in general, what properties exist, but the task of
the natural scientist. In principle, ‘sparse’ properties should be the minimum basis
sufficient to fully characterize the world (see Lewis 1983, 12) or, in linguistic terms,
they should be what primitive predicates refer to in a language sufficient for a
complete description of everything. In the following, these sparse properties will
be referred to as “authentic properties”.11
Yet although it has been usual to maintain that it is natural science that determines
a posteriori, by empirical research, what properties exist, it should not be assumed
that the postulation of authentic properties is connected with some form of natural-
ism. Several advocates of theories of authentic properties, such as David
M. Armstrong (cf., for example, 1997, 5–10), are naturalists, that is, they hold that
everything that exists is the space-time system and the entities it contains that are
described by natural science. But there is no systematic connection between natu-
ralism and the postulation of authentic properties. The claim is simply that not every
10
An infinite plurality is said to be ‘denumerable’ if it can be put in bijection with the set of natural
numbers, as with the set of rational numbers or the set of integers. An infinite plurality is said to be
‘indenumerable’ if it cannot be put into bijection with the set of natural numbers, just as happens
with the set of real numbers.
11
This is not to deny, of course, that there is an important philosophical problem about what is the
content of our thinking, and what semantic values should be given to the expressions of our
languages. Nothing prevents, in particular, that operations from universals to universals should
allow the generation of such contents—as has been supposed by those who have defended
‘abundant’ properties—from a base constituted only by authentic properties, according to what is
indicated here. These, however, are further issues that will not be discussed in this work.
8 1 Introduction
term or predicate of our language must be correlated with some authentic property
that has to be its ‘meaning’, and that it is more reasonable to think that there are real
properties for which we do not have linguistic expressions that designate them.
Properties are not the mere correlates of our languages. Moreover, according to
everything we know, there could be properties of such a character that they will
evade our best efforts to discover or understand them. These properties could still be
the kinds of properties with which natural science deals. Something like this is what
we should suppose if everything that exists is of a physical nature, but it is not what
we should suppose if there are entities that transcend time and space. There could be
properties of a non-physical nature. There may be mental properties, for example,
that are not reducible to physical properties or not based on physical properties. If so,
such mental properties need to be entered into a complete characterization of the
world.
§ 3. The expression “possible world” will be used freely. As is well known, there is
controversy about what should be understood by this term (cf. for a general per-
spective, Divers 2002; Alvarado 2008). Despite the differences in perspective, there
is a minimum content that can be assigned to the expression, and which has utility
when it comes to considering modal facts—that is, facts about what might be the
case, or about what necessarily is the case, or about what is not necessarily the case,
etc. By ‘possible world’ we should understand a form in which all the things could
be. Anyone who believes that things could be different from how they actually are
believes that there are ways in which things could be that is not the way things are
(see Lewis 1973, 84). Accepting the existence of possible worlds, therefore, is little
more than accepting the existence of possibilities, which seems a matter of common
sense. No special assumption is made here about their nature.
Modal issues are considered in several sections of the work (see §§ 24, 30,
41–44). It will be seen that when it comes to an understanding of the nature of
worlds, an ‘actualist’ conception will be preferred, according to which there is only
one concrete world, the actual world. ‘Possible worlds’ are abstract constructions
that represent how things could be. For the moment, however, it will not be
necessary to assume anything more than that to speak of ‘possible worlds’ is simply
to speak of metaphysical possibilities considered globally.
§ 4. Much of the argumentation in this chapter will have to do with the aptitude or
inaptitude of certain postulated entities to ‘explain’ specific facts. The notion of
‘explanation’, however, has been the subject of much controversy throughout the
past century. An important tradition has connected it with the notion of ‘causality’.
This connection to causality makes its application in ontology difficult, since it is
doubtful, for example, that universals ‘cause’ the phenomena of one over many (cf. §
10). An ‘ontological’ explanation must have a different character from what is
expected of an explanation in the natural sciences. In a very general sense, one
1.2 Possible Worlds, Grounding, Dependence 9
12
The reason why it has usually been assumed that grounding is a relationship that has as relata
‘facts’ is that many analyses have focused on the uses of the expression ‘–because–’ and the like,
that connect complete sentences. For this reason, it has also been proposed that it be treated as a
sentential connective (see Fine 2012). If someone has difficulties with the freedom with which the
relations of grounding between entities of any category will be treated here, one can replace the
expressions ‘x is grounded on y’ with ‘the fact that x exists is grounded in the fact that y exists’. Of
course, this is not the place to make a detailed discussion of this topic. For a recent defence of
‘entity’ grounding, see Wilhelm, 2019.
13
With a better understanding of the nature of grounding, other hypotheses have been explored with
infinite or cyclic structures of grounding in which there is nothing ‘fundamental’, not based on
anything (cf. for example, Bliss 2013, 2014). Although none of these hypotheses is inconsistent, it
is far from clear that there are cases in which it is reasonable to postulate such structures. Here we
are going to assume that both grounding and dependence are ‘well-founded’ relationships, so there
must be something fundamental, not grounded on anything, or independent, not dependent on
anything.
10 1 Introduction
distinction, intrinsic properties are those that are not relational. None of these
strategies has been entirely satisfactory (see Weatherson and Marshall 2012). Of
course, it is not possible to adjudicate this question here. Probably, there is no single
precise concept of ‘intrinsic property’. There are certain relatively vague intuitions
about properties that determine an object regardless of what other objects exist and
what relationships exist with them, but these intuitions are consistent with different
and incompatible ways of making them precise. It seems more prudent, for these
reasons, to define two different concepts of ‘intrinsic property’ according to two
different criteria. The first is the one that corresponds to combinatorial modal
intuitions:
[Combinatorially Intrinsic] P is combinatorially intrinsic ¼ df the fact that
x instantiates P is independent of whether x is alone
or accompanied
Of course, a ‘combinatorially extrinsic’ property is one that is not ‘combinatorially
intrinsic’. This formulation comes from Lewis and Langton (1998).14 An object x is
alone in the possible world w if and only if there is no other object than x in w. An
object x is accompanied in w if and only if it is not alone in w. According to this first
way of characterizing what an intrinsic property is, it is one whose instantiations do
not vary if the possible world in question includes more or less different objects. If
something has 5 gr of mass, for example, its mass does not vary if there is another
object or not. On the contrary, the property of being 5 metres from a cube is not
intrinsic because nothing will instantiate it in possible worlds in which there is no
cube. Adding or removing a cube makes a difference to whether something has the
property of being 5 metres from a cube. Note that if there are objects of necessary
existence, these will not vary between different possible worlds, because no world
lacks them. If the number 3 is a necessary object, then the property of being
accompanied by the number 3 will be combinatorially intrinsic to every object,
since the fact that an object is accompanied by the number 3 is invariant concerning
worlds in which that object is or is not accompanied by others. For this property to be
combinatorially extrinsic, there must be possible worlds in which adding or remov-
ing an object to a possible world determines that something is no longer accompa-
nied by the number 3 or that it becomes accompanied by the number 3. However, as
the number 3 is—ex hypothesi—modally invariant, nothing that is added or removed
modally produces a difference.
14
There are sophistications introduced in the analysis of Lewis and Langton that are not important
here. These sophistications are motivated by the possibility of disjunctive properties. The proposed
analysis ultimately depends on a primitive distinction between ‘natural’ properties and those that are
not (see Lewis and Langton 1998, 118–121). A property is said to be basic intrinsic if and only if:
(i) the fact that an object instantiates it is independent of whether the object is alone or accompanied
(as indicated above), (ii) it is not the disjunction of other natural properties, and (iii) it is not the
negation of a natural disjunctive property. Two objects are duplicates if and only if they have the
same ‘basic’ intrinsic properties. Now, a property is intrinsic if and only if it cannot differ between
duplicates.
1.3 Intrinsic and Extrinsic Properties, Mereology, Concrete and Abstract 13
15
Although cf. Lewis 1983, 26, note 16; Lewis calls ‘external’ the dyadic relations that are
supervening on the pair of relata considered together, although they are not supervening on the
intrinsic natures of these relata.
16
The expression ‘internal relationship’ has also been used to designate a relationship that is
essential to their relata. That is, if R is an internal relation for x1, x2, . . ., xn, then none of those
objects would exist if it were not in the relation R with the rest. The designation ‘internal’ has also
14 1 Introduction
been used for a relationship that has a plurality of objects that are not entirely independent of each
other. These notions will not be considered in this work.
17
The most important difficulties that have arisen for these proposals of analysis have to do with the
status that should be granted to entities that have as ‘constituent’ something that is clearly concrete.
Attempts have been made to distinguish between the concrete and the abstract through, for example,
the distinction between what has spatio-temporal location and what does not, what can enter into
causal interactions and what cannot, what is necessary and what is contingent, and also in relation to
features of categories of entities. Thus, Hoffman and Rosenkrantz (1994, 182–187) argue that
something is concrete if and only if it instantiates a category C such that C possibly has at least one
instance with spatial or temporal parts. Assume the particular object b, which is a clear case of a
concrete entity. It is, in effect, localized spatio-temporarily, it can enter into causal interactions, it is
contingent, and it belongs to a general category of entities that possibly possess instances with
spatial or temporal parts. Consider now the set {b}. Is it located where b is located? Does it
1.4 Summary of What Is to Come and some Nomenclature 15
will also be assumed that the distinction satisfies certain theoretical restrictions
(cf. Cowling 2017, 70–71): (i) something is concrete if and only if it is not abstract,
and is abstract if and only if it is not concrete (although cf. Williamson 2013, 7);
(ii) everything is either concrete or abstract; (iii) if something is concrete, then it is
necessarily concrete, and if something is abstract, then it is necessarily abstract; (iv) a
mereological sum, one of whose parts is concrete, is a concrete entity.
§ 8 The central thesis defended in this work is that, at the fundamental level, all that
there is are universals and their instantiations. To defend this thesis, it is necessary to
justify, in the first place, the claim that there are universals. This is done in Part I
(Chaps. 2, 3, 4, and 5, §§ 9–38). As is well known, the philosophical discussion
around universals has undergone a profound transformation in the last fifty years,
especially since the work of Armstrong (1978a, b, 1983, 1989, 1997). It had been
traditional to approach the question about the existence of universals as a problem
about the ontological commitments to which our everyday communicative practices
compel us, in which it seems that we quantify over universals, or reference is made to
universals (see on this approach, Jackson 1977). After Armstrong’s contributions,
however, the postulation of universals has come to be seen as having more to do with
the roles that must be fulfilled for the explanation of objective resemblances—as
these are described in our best scientific theories—as well as causal connections and
natural laws. This is the approach followed here. If the reasons for the postulation of
universals as part of our ontology have to do with specific theoretical roles that
cannot be satisfied by other entities, then it will be essential to pay some attention to
the nature of such ‘roles’. It will also be essential to carefully consider whether
universals systematically fulfil these roles better than their alternatives. Chap. 2 (§§
9–15) explains the theoretical roles usually attributed to universal properties or their
systematic alternatives. In Chaps. 3, 4, and 5 (§§ 16–38), a comparative evaluation is
made between universals and those alternatives. As will be explained later, this
relative weighting will not be carried out with respect to all the conceptually possible
alternatives but only with respect to the conceptions that seem more relevant, at least
for me. These are: resemblance nominalism (Chap. 3, §§ 16–24), the theories of
tropes (Chap. 4, §§ 25–31), and theological nominalism (Chap. 5, §§ 32–38). The
question that will be considered in Part I, then, is whether there are universals rather
than resemblance classes of objects, resemblance classes of tropes, or concepts in the
mind of God.
intervene in the causal relationships in which b intervenes? Since, in general, a set depends on its
elements, the singleton set {b} is contingent, since it does not exist if b does not. It is not obvious if
{b} should count as ‘concrete’ after all. Similar considerations could be made regarding the
proposition b exists, and of the property of being identical to b.
16 1 Introduction
The justification of the claim that there are universals, however, leaves open
whether universals are immanent or transcendent. This is discussed in Part II
(Chaps. 6, 7, 8, 9, and 10, §§ 39–77). Traditionally, those who have postulated the
existence of transcendent universals have basically done so through considering the
functions that universals would have as the content of thought and language. Since
here it is assumed that the properties are those that are required to fulfil the
theoretical roles indicated in Chap. 2, it cannot be assumed that for each concept
that is the content of our judgments there must be a correlative property. The way in
which the existence of transcendent universals will be justified—or, more precisely,
the conditional according to which, if something is a universal, it must be a
transcendent universal—will be through considering the theoretical functions of
universals in three areas: modal metaphysics (Chap. 6, §§ 39–44), the metaphysics
of natural laws (Chap. 7, §§ 45–52), and the profile of ontological priority that a
universal must have (Chap. 8, §§ 53–59). In each of these cases, it will be shown that
the only alternative coherent with the functions attributed to a universal is to suppose
that it is a universal whose existence does not depend on having any instantiation.
For many, however, these arguments will not be persuasive if the difficulties usually
directed against them are not addressed. In Chap. 9 (§§ 60–74), then, such difficul-
ties are considered, and I show that none of them is a reason to reject transcendent
universals. Attention will be focused on three principal objections: the assumption
that a transcendent universal would have no causal power, nor would make any
difference in the causal powers of something; the criticism that the postulation of
transcendent universals would be less economical ontologically than the postulation
of immanent universals; and the epistemological objection that transcendent univer-
sals would not be cognizable by our ordinary cognitive capacities. Part II includes a
chapter devoted to considering the question of what are the conditions of identity for
transcendent universals (Chap. 10, §§ 75–77). It goes on to show that, contrary to
what might have been assumed, transcendent universals make up a unitary structure
of necessary existence. The universals are the nodes of such a structure.
The postulation of transcendent universals leaves unaddressed the question about
how particular objects should be understood. This is dealt with in Part III (Chaps. 11,
12, and 13, §§ 78–94). Given what has been argued, especially in §§ 67–70, the
ontology of particular objects must include tropes that are essentially the particular
instantiation of a transcendent universal. In the abstract, it would be consistent with
the postulation of tropes with these characteristics to understand particular objects
with a substratum that is characterized by tropes but also to understand particular
objects as trope bundles. In Chap. 11 (§§ 78–81), these theoretical alternatives will
be presented, along with the traditional difficulties directed against ontologies of
substrata and ontologies of bundles. One of the alternatives seems especially
appropriate for solving the problems discussed in Chap. 11: the so-called ‘nuclear
theory of bundles of tropes’ (Chap. 12, §§ 82–86). This is a form of a theory of trope
bundles, but with two important features that distinguish it from theories of tradi-
tional trope bundles. In this conception, the unity of an object is grounded on
ontological dependence relations, and there is a distinction between essential and
accidental properties for an object. This is achieved with the distinction between two
1.4 Summary of What Is to Come and some Nomenclature 17
different ontological strata in a particular object, called the “core” or “nucleus” and
the “periphery”. Despite the advantages offered by the traditional theory of nuclear
trope bundles, it has serious problems that have to do crucially with how the relation
of ontological dependence must be understood (especially § 86). This makes it
convenient to propose a ‘reformed’ nuclear theory (Chap. 13, §§ 87–94) that evades
these difficulties by postulating a nucleus consisting of a single trope. The reformed
nuclear theory retains the advantages of the traditional nuclear theory, but also
achieves the unification of the ontologies of properties and substrata, customarily
considered as incompatible with each other. The only nuclear trope is, in effect, a
substratum (see § 89). The reformed nuclear theory is a theory that conceives
particular objects, at the same time, as bundles of tropes and as constituted by a
substratum. We also consider how the reformed nuclear theory can make intelligible
sense of structures formed from beginning to end only by relations, without relata
(cf. §§ 93–94). Ontologies of this type have been of interest to conceive a multitude
of different types of physical entities, both at the quantum level and for the grand
scales of the universe. The reformed nuclear theory has enough flexibility to give an
adequate treatment of these structures.
I have tried to offer in this book a comprehensive defence of a Platonistic
ontology of universals that presents the advantages and discusses its counter-
intuitive consequences. In effect, for an adequate assessment of the feasibility of
such an ontology it is necessary to have a synoptic perspective of those theoretical
advantages and disadvantages. I do not intend, then, to convince the reader just with
some short argument, but with a much more arduous revision of most of the reasons
that have been put forward in decades of philosophical discussion –in some cases,
centuries of philosophical discussion. So, it has been convenient to include here the
exposition of many ideas that for most specialists are hardly new, to gain a broader
perspective appropriate for a correct theoretical assessment. Besides, a more com-
prehensive presentation of the range of questions that involve universals has the
additional advantage that the book results more accessible to non-specialist readers
that want to be introduced to the ontological debates that is both informative and
‘opinionated’.
Throughout this work, reference will be made to predicates (that is, sequences of
phonemes or graphemes of some language), concepts (that is, the content of mental
states in which we judge or consider something), and universal properties. To avoid
confusion, predicates will be presented in double quotes, as usual. Concepts, on the
other hand, will be presented in single quotes. A universal property, however, will be
presented in italics. In this way, “__ is a cube” is the predicate that is truthfully
attributed to something that is a cube; ‘being a cube’ is the concept that a rational
subject truly judges of something that is a cube; being a cube, on the other hand, is
the universal property that all objects that are cubes instantiate. Analogously, the
propositions—that is, the content of what is stated by whoever states a complete
18 1 Introduction
References
Alvarado, J. T. (2008). ¿Qué son los mundos posibles? Intus legere, 2(1), 1–23.
Alvarado, J. T. (2013). Fundación y reducción. Aporía. Revista internacional de investigaciones
filosóficas, 6, 59–74.
Armstrong, D. M. (1978a). Universals and scientific realism, Volume I, Nominalism and realism.
Cambridge: Cambridge University Press.
Armstrong, D. M. (1978b). Universals and scientific realism, Volume II, A theory of universals.
Cambridge: Cambridge University Press.
Armstrong, D. M. (1983). What is a law of nature? Cambridge: Cambridge University Press.
Armstrong, D. M. (1989). Universals. An opinionated introduction. Boulder: Westview.
Armstrong, D. M. (1997). A world of states of affairs. Cambridge: Cambridge University Press.
Barnes, E. (2012). Emergence and fundamentality. Mind, 121, 873–901.
Bliss, R. (2013). Viciousness and the structures of reality. Philosophical Studies, 166, 399–418.
Bliss, R. (2014). Viciousness and circles of ground. Metaphilosophy, 42, 245–256.
Correia, F., & Schnieder, B. (2012). Grounding: An opinionated introduction. In Fabrice Correia &
Benjamin Schnieder (eds.), Metaphysical Grounding. Understanding the Structure of Reality.
Cambridge: Cambridge University Press, pp. 1–36.
Cowling, S. (2017). Abstract entities. London: Routledge.
Divers, J. (2002). Possible worlds. London: Routledge.
Ehring, D. (2011). Tropes: Properties, objects, and mental causation. Oxford: Oxford University
Press.
Fine, K. (2012). Guide to ground. In Fabrice Correia & Benjamin Schnieder (eds.), Metaphysical
Grounding. Understanding the Structure of Reality. Cambridge: Cambridge University Press,
pp. 37–80.
French, S., & Krause, D. (2006). Identity in physics: A historical, philosophical, and formal
analysis. Oxford: Clarendon Press.
Goodman, N., & Quine, W. V. O. (1947). Steps towards a constructive nominalism. The Journal of
Symbolic Logic, 12, 105–122.
Hoffman, J., & Rosenkrantz, G. (1994). Substance among other categories. Cambridge: Cambridge
University Press.
Jackson, F. (1977). Statements about universals. Mind, 86, 427–429.
Lewis, D. K. (1973). Counterfactuals. Oxford: Blackwell.
Lewis, D. K. (1983). New work for a theory of universals. Australasian Journal of Philosophy, 61,
343–377. Reprinted in David Lewis, Papers in Metaphysics and Epistemology. Cambridge:
Cambridge University Press, 1999, pp. 8–55.
Lewis, F. A. (2013). How Aristotle gets by in Metaphysics zeta. Oxford: Oxford University Press.
Lewis, D. K., & Langton, R. (1998). Defining ‘intrinsic’. Philosophy and Phenomenological
Research, 58, 333–345. Reprinted in David Lewis, Papers in Metaphysics and Epistemology.
Cambridge: Cambridge University Press, 1999, pp. 116–132.
18
In general, I have tried not to include any formulations in the usual symbolism of formal logic.
When this has seemed useful for greater precision, expressions in logical symbolism have been
included in footnotes. Quantified first-order modal logic is used as a framework for these formu-
lations, without Barcan’s Formula or the Converse Barcan’s Formula.
References 19
By this time our hero had become acquainted with Dr. Oudney, at
whose house we have had the pleasure of occasionally meeting him;
and when the Doctor was appointed to his exploratory expedition to
Africa, he expressed, through the medium of the common friend of
both, and to whose information we have been much indebted in
drawing up our memoir of their lives, his desire to be attached to the
mission. Clapperton could not boast the possession of much either
of the literary or the scientific knowledge requisite to constitute the
intelligent traveller; but he was distinguished for other qualities fitted
to render him a valuable acquisition to any mission similar to that to
the accomplishment of which Dr. Oudney had been appointed. The
portrait prefixed to the “Journal of his second Expedition,” shows that
his figure was tall, strong, and manly. He had a fine bust, and his
whole frame combined length of arm, great strength, weight, and
agility—circumstances which the portrait does not sufficiently
represent, and is also deficient in expressing his fine lion-like
forehead and eye. We have seen that he was endowed with a
constitution of almost invincible strength, that he possessed a most
enterprising disposition of mind, great conscientiousness in the
discharge of duty, and a heart alive to the kindly impressions of
compassion, and capable of strong and steady friendship. Such a
travelling companion was likely to be a treasure to a man like Dr.
Oudney; and he had the pleasure to be informed that his application
to have Clapperton attached to the mission was granted.
Accordingly, in the autumn of 1821, the travellers left Scotland for
London, with the view of then commencing their expedition to the
interior of Africa. In a letter to a friend dated London, September 1,
1821, Clapperton says, he had been supplied with arms, and had got
instruments of his own choosing, and mentions the sextant as the
most complete he had ever seen; he states to his friend that he had
had several agreeable interviews with his uncle; and adds, that he
was just on the eve of setting off for Falmouth. His next letter to the
same gentleman was written at Mourzuk, May 20, 1822, in which he
tells his friend that his health had continued vigorous, although the
heat was 106 degrees of Fahrenheit, in the shade; and says, that
Oudney was much admired by the ladies for the blackness of his
beard, and himself for the strength of his mustachoes. Oudney in a
postscript on the same sheet, says, “Clapperton is just the old man.
He is a strange-looking figure with his long sandy coloured beard
and mustachoes. You would smile were you to see him smoking his
pipe, and calling to his servant, Waddy ama simpri, or fill my pipe.” In
a subsequent letter from the same place, to the same
correspondent, Clapperton speaks in praise of the Tuaricks, whom
by this time, (Sept. 1822) he had visited. He says they are a fine
warlike race, who fear nothing but the devil and his agents, that they
offered to convey both him and Oudney to Timbuctoo; and adds,
“They wished me much to take a wife amongst them, but I said she
would have to go to Bournou and England with me, which got me out
of the scrape with a good grace, as their women never leave their
country, and those who marry them must stay with them.” And the
fact is that our hero very soon found himself as much at home
among the wild Tauricks, who traverse the sandy deserts of
Sahaara, as he had formerly done among the Indians who dwell in
the midst of the forests of Canada.
It would seem that Clapperton did not regard it as any part of his
duty to keep a separate journal while Oudney lived; nor was it
necessary, as they were generally together in all the excursions
which they made in Fezzan, and their joint observations were
combined by the Doctor into the same narrative, to which he put his
own name. But the case was greatly altered after the arrival of the
travellers in Bournou, where Oudney was seized with the illness
which terminated in death, upon the 12th of January, 1824. After this
mournful event, Clapperton, sick and sorrowful as he was,
proceeded onward to Kano, with the view of visiting Sackatoo, as
was originally intended. He reached this city, (as may be seen in his
printed journal) upon the 16th of March, and had many interviews
and long conversations with the sultan, Bello. He remained at
Sackatoo till the 4th of May, when he began to retrace his steps,—
again reached Kuka upon the 8th of July, and arrived in London in
the summer of 1825. Clapperton and Denham came from Tripoli to
Leghorn, sent the animals and baggage home by sea, under the
charge of Hillman, their only surviving companion, while they
themselves crossed the Alps, and on the 1st of June, 1825, they
reported their arrival in England to Earl Bathurst, under whose
auspices the mission had been sent to Africa.