0% found this document useful (0 votes)

47 views

1 Combinatorial Predication and The Ontology of Unit Attributes

Uploaded by

jtpaasch

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

47 views

1 Combinatorial Predication and The Ontology of Unit Attributes

Uploaded by

jtpaasch

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 42

1

Combinatorial Predication
and the Ontology of Unit Attributes*

The topic of relations may prove a peculiarly rewarding

one for the theory of universals as a whole. For it may be
expected that important classifications and distinctions
which can be made at the level of polyadic universals will
vanish, or become merely notional, in the limiting monadic
case.
D. M. Armstrong
A Theory of Universals

1. Introduction

A primary task of ontology is to provide a satisfactory theory of the ulti-

mate categories of entities and the relationships between them. It is no
exaggeration to say that the majority of the issues and distinctions proposed
in this regard are tied, directly or indirectly, to what would be answers to a
single, intuitive, and simply stated problem. This is the Triple Aspect Prob-
lem:

What coherent account is there for the ubiquitous, fundamental epis-

temic given of apparently unrepeatable entities (individuals, particu-
lars) possessing (instantiating, having as ontic (‘material’) predi-
cates) apparently repeatable attributes (universals, intensions)?

Central here is the pre-critical unrepeatability/possession/re-peatability tri-

chotomy everywhere exemplified in our internal and external experience.

Apple a, for example, is apparently an unrepeatable individual possessing
as predicates a number of repeatable properties, e.g., Red, Round, Has-
mass. Likewise for relations, e.g., it is intuitive that the dyadic Square-root
relation is numerically identical among distinct relata pairs: <2,4>, <3,9>,
<4,16>, etc. Independent of whether natural numbers are universals or par-

Mertz, D. W. (2015). Essays on realist instance ontology and its logic. ProQuest Ebook Central <a onclick=window.open('http://ebookcentral.proquest.com','_blank')
href='http://ebookcentral.proquest.com' target='_blank' style='cursor: pointer;'>http://ebookcentral.proquest.com</a>
Created from georgetown on 2021-01-03 10:12:16.
2

ticulars,1 pre-critically each relata pair jointly possesses as a shared ontic

predicate the same Square-Root relation.

Of the three thus identified ontic aspects of individuality, predicabil-

ity, and universality, predication is epistemically primary. For, there is no
explication of either the unrepeatable aspect (the non-recurrent ‘thisness’
(haecceitas) of individuals) or the repeatable aspect (the recurrent ‘such-
ness’ (quiddity, natura) of universals) without an appeal to the ‘posses-
sion’ of predication. Aristotle described this pivotal role saying: “Some
things are universal, others individual. By the term ‘universal’ I mean that
which is of such a nature as to be predicated of many subjects, by ‘individ-
ual’ that which is not thus predicated.” (De Interpretatione 17a38-40; cf.
Metaphysics 1038b10) Accordingly, an entity that is not predicable, or at
least not predicable of multiple subjects, is an unrepeatable individual.2
Despite its fundamental and linking role in the order of explanation, the
parallel role of predication in the order of reality has gone generally under-
appreciated, this due in large part to the traditional and misleading con-
tainment (or inherence) model of prediction. In the dialectic that is the his-
tory of ontology, ontic predication has been given a mostly secondary and
derivative role, theories of predication following whatever were theorized
as the more fundamental natures3 of individuals and/or universals. Indica-
tive is the fact that historically one finds the Triple Aspect Problem limited

∗
Original version first published in The Modern Schoolman LXXIX (2002): 163-97.
1
I have argued that natural numbers are best described as both particulars and univer-
sals in the following sense. As predicable of concepts, e.g., in the adjectival ‘There are
three apples on the table’, numbers are individuated instances predicated of concepts,
(∃3i)3i(Concept of apple-on-the-table), and in abstraction from its instances a number
is a universal, e.g., 3 is the universal content common to its instances 3i, 3j, 3k,…, and
is referred to in substantival number sentences such as ‘Three is a prime number.’ The
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

advantages of this view are given in D. W. Mertz, Moderate Realism and Its Logic (New
Haven: Yale University Press, 1996), pp. 259ff.
2
It is to be noted that Aristotle’s criterion as stated here does not require that an indi-
vidual be a non-predicable entity. And indeed, in the realist version of instance ontol-
ogy advocated herein (and for which there is evidence in Aristotle in regard to proper-
ties—see Moderate Realism, pp. 98-111) it is individual relation (including property)
instances Rni that are the only predicable entities.
3
The term ‘nature’ is used herein as equivalent to essence or quiddity (‘whatness’) and
as meaning that which is described in a definition.

to the Problem of Individuation,4 and/or the Problem of Universals,5 but

there is no comparably explicit ‘Problem of Predication’. The closest to
such an identified problem is what is now called Bradley’s Regress, an an-
cient and perennial argument from which have been drawn contradictory
conclusions. It is a thesis herein that this ambiguity is mutual with the sec-
ondary role given predication, and that both are the consequence of a fail-
ure on the part of ontology to give an adequate analysis of ontic predica-
tion’s combinatorial6 nature and to draw what are its significant implica-
tions.

Using polyadic relations as a means of epistemic ‘magnification’, it

will be argued herein that all ontic predicates insofar as they are predicable
of their subjects are properly and succinctly characterized as individuated
states of intensioned connectivity. Among the consequences of this analysis
of ontic predication is, first, that it provides a solution to the Problem of
Individuation that avoids the inadequacies of current theories. These theo-
ries are principally: a) the posit of what must be absolutely contentless or
‘bare’ individuators (Armstrong, Moreland, Bergmann)7, b) the individua-

4
The Problem of Individuation can be stated as: What explanation can be given of the
fact that, though absolutely every characteristic of an individual is a repeatable inten-
sion, that entity can nevertheless be unrepeatable qua individual?
5
The Problem of Universals can be stated as: What explanation can be given of the
apparent fact that numerically one and the same intension can be predicably ‘in’ di-
verse subjects—unum in multis?
6
Cognizant of its mathematical usage, I have nevertheless chosen the term ‘combina-
torial’ as appropriate in a literal sense since on the following analysis predicates are
‘combinators’, i.e., agents of plural unification.
7
D. M. Armstrong, A World of States of Affairs (Cambridge: Cambridge University Press,
1997), pp. 68, 109; James P. Moreland, ‘Theories of individuation: A Reconsideration of
Bare Particulars’, Pacific Philosophical Quarterly 79 (1998): 251-63, and Universals,
Qualities, and Quality-Instances: A Defense of Realism (Lanham: University Press of Amer-
ica, 1985); Gustav Bergmann, Logic and Reality (Madison: University of Wisconsin Press,
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

1964), pp. 133-34, 277. In regard to Armstrong, he refers to bare particulars as ‘thin par-
ticulars’. By this designation he seeks to emphasize his thesis that in extra-conceptual real-
ity there are no particulars without some property attached as an ontic predicate, i.e., there
are only states-of-affairs or facts. A thin (bare) particular is a conceptual abstraction from
a state-of-affairs, as is the property or relation universal, the only other constituent of a
state-of-affairs. Also see Armstrong’s Nominalism & Realism: Universals & Scientific Re-
alism, Vol. I (Cambridge: Cambridge University Press, 1978), pp. 113ff. The point I wish to
emphasize is that Armstrong’s thin particulars are for him fundamental and distinct onto-
logical entities, inseparable as they may be from universals, and that they are themselves

tion-by-instantiation of especially empowered substance universals (Loux,

Lowe)8, or c) the assigning to individuation the status of an unexplainable
primitive (a nominalist tack, e.g., Ockham, Campbell)9. Contrary to these
theories that take individuation to be either primitive or the posited attrib-
ute of a posited primitive entity, on the following individuation will be
seen to follow logically from what must be the nature of ontic predication.
Secondly, the combinatorial assay of predication contributes to solving the
Problem of Universals by explaining the sense in which intension univer-
sals are ‘in’ and yet, as necessitated on the following, not ontically predi-
cable of the individuals they characterize. For the problem with universals
is not simply that numerically the same entity is ‘in’ as a constituent of
each of distinct complex wholes, this type of unum in multis being a com-
mon-place, e.g., a in {a,b} and {a,c}, or as a part that is shared by multiple
machines. Assertions to the contrary may derive from what is a thesis of
naturalism that for an entity to exist it must necessarily be a relatum of spa-
tial relations, this on the relationist analysis of space asserted here,10 and

bare of any content or intension—“…The thin particular, the particular apart from its prop-
erties” (States of Affairs, p. 115).
8
Michael Loux, Substance and Attribute (Dordrecht: Reidel, 1978), pp. 163ff. Also, Loux,
Metaphysics: A Contemporary Introduction (New York: Routledge, 1998), pp. 117ff.; E. J.
Lowe, The Possibility of Metaphysics (Oxford: Clarendon Press, 1998), pp. 180-83, 197.
9
Indeed, some nominalists take the line that individuation is a non-problem. In the
words of Ockham, “One does not have to look for a cause of individuation…. Rather
one has to look for the cause why it is possible for something to be common and uni-
versal.” Paul Spade, Five Texts on the Mediaeval Problem of Universals (Indianapolis:
Hackett Publishing Co., 1994), p. 172. Keith Campbell takes this line in Abstract Par-
ticulars (Oxford: Basil Blackwell, 1990), p. 69.
10
Adopted here is the relationist over the substantivist view of space. For a justifica-
tion of this view see Moderate Realism, pp. 39-42. There it is argued that the substan-
tivist view that space(-time) is composed of points(-instances) fails because the latter
reduce upon analysis to what would be bare particulars reserved as relata for spatial(-
temporal) relations. Apparently these bare point-instances would be the same discrete
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

space-time minima essential to the geometrodynamics conception of real space-time

Campbell has advocated, e.g., in ‘Unit Properties, Relations, and Spatio-Temporal
Naturalism’, The Modern Schoolman LXXIX (2002): 151-62. Further, it is instructive
to note that Campbell holds space to be real as demonstrated by its property of right- &
left-handedness (enantiomorphism) and hence concludes that a Leibnizian relationist
conception of space is false, since the latter includes a characteristic foundationism (a
view advocated by Campbell himself) claiming the eliminative reduction of relations
to properties of their relata. The point I would make is that relationism need not be
Leibnizian. The counter-theses advocated here are that foundationism is false, that re-

the fact that the same entity can not be in two distinct places at once. Asso-
ciated naturalist theses are that the existence of spatial(-temporal) relations
is prerequisite to, or at least concomitant with, the existence of causal rela-
tions, and that ‘There is no entity without causal efficacy’, causality here
being efficient causality.11 Yet, the spatial criterion for existence is contra-
dicted by irreducible categories of entities that do not support spatial (or
causal) relations as relata, e.g., concepts, propositions, logical operators,
sets, numbers; at least some instances of the laws of logic, mathematics,
and science; and even instances of spatial relations themselves (e.g., try to
cogently specify the spatial location of the Taller-than relation when per-
son a is taller than person b). Any attempt to limit the real to the concrete,
even with space conceived substantivally as a plenum of geometric points,
will be defeated by the necessity of appealing to abstract relations, e.g., the
relation Located-at. For, under this point-plenum conception of space, the
reality of an entity a is a function in part of its being located at some set or
other of spatial points, s, at a temporal instance. That is, for what are non-
identical a and s, it is the reality of the further and necessary asymmetric
Located-at relation existing between them that is essential to conveying an
existential status from s to a—a’s existence ontically presupposes both that
of s and its being Located-at s. But then to have the requisite reality by the
same criterion, the Located-at relation between a and s itself ontically pre-
supposes being located, presumably at s, and as located in space this rela-
tion between a and s would be individuated, i.e., be a unit attribute or in-
stance. Yet the locating of the latter instance would itself be a further onti-
cally presupposed instance of the Located-at relation existing between the
first instance and s, and clearly this is the beginning of a vicious regress.
The reality of every relation instance in this chain presupposes the reality
of a prior one. Consequently, the instance of the Located-at relation exist-
ing between a and s cannot be located, i.e., it is not concrete, and so the ex-
clusive spatial criterion for existence must be rejected.

Further and relatedly, under the spatial criterion the abstract entities
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

and relations of, say, logic and mathematics would have to reduce to either
illusion or to a mysterious ‘supervenient’ state somehow between being
and non-being and inherited from, presumably, the spatial locatability of
brain states. But then in regard to the latter alternative a regress analogous

lations and their instances are fully real and unreducible, and that space(-time) is con-
stituted by instances of space(-time) relations.
11
Armstrong, States of Affairs, pp. 5-6, 41-43.

to the one above would be evoked by the necessitated reality of the Super-
venience relation, and this in addition to the Located-at regress in regard to
brain states. Moreover, here it would seem one cannot avoid concluding
that, for example, natural numbers, if they have any reality at all, must in-
herit the status of entering into spatial relations, and this is counterfactual.
It is not the case that 2 is spatially related to 3 anymore than spatial point x
is numerically (or causally, legally, ethically, emotionally, etc.) related to
spatial point y. Yet, none of this in itself implies anything about the reality
or unreality of numbers, spatial points, or any of the relations mentioned.
The ontically privileged role reserved for relations with spatial intensions
is as much of an exaggeration (and limitation) as the Pythagorean thesis
that what is real is marked ultimately by being relata for abstract relations
with numerical intensions. The naturalist’s only alternative for countering
these objections is to argue that relations (and their instances), particularly
asymmetric relations, can be eliminated from one’s ontology by reducing
them to property instances of their spatially located relata. But this is an
impossibility as maintained below.

Setting the spatial bias aside, realists have pointed out correctly that
there is nothing inherently incoherent in the notion of numerically the same
entity being a constituent of diverse wholes, and this whether the latter en-
ter into spatial relations or not. Rather, ‘the problem’ with universals is at
bottom, I propose, a problem concerning ontic predication, one that derives
from the classic mis-identification of universals with ontic predicates. This
mis-identification results in the puzzle of how a monadic property whose
intension specifies a predicable nexus with exactly one subject can, as a
universal, be simultaneously predicable of multiple subjects. The puzzle is,
of course, repeated for predicates of every n-adicity, though it is monadic
properties that have been the traditional paradigm. And, it is precisely this
latter ‘tyranny of the monadic’ that has served to mask what is a corrective
more obvious in the case of polyadic relations. For, given the unreducibil-
ity of the polyadic to the monadic claimed herein, the intension of an n-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

adic relation (e.g., Taller-than) requires a simultaneous predicability among

multiple, i.e., n, subjects. Hence, the puzzle over predication is not simply
that one and the same predicate can have multiple subjects, but is rather
how numerically the same n-adic predicate, that as such specifies exactly n
subjects, can be simultaneously predicable of (repeatable over) an indefi-
nite number (not necessarily n) of n-tuples. A now obvious answer to the
puzzle is that n-adic predicates, including the monadic as the limiting case,

are in fact not repeatable, i.e., are individuated to specific n-tuples. That
this is indeed the solution to the Problem of Universals will be seen to fol-
low from the formulation and solution of the more basic Problem of Predi-
cation. Using relations as a perspicuous medium for analysis we shall see
that ontic predicates themselves are not repeatable and they are not ‘in’,
i.e., not internal constituents of, their subjects, though the predicates’ char-
acterizing intensions are ‘in’ them as constituents. That is, an intension can
be a non-predicable constituent of each of multiple predicates, but the sub-
suming predicates are neither universal nor in their subjects. The result is a
moderate realism absent the obscurities found in the versions of Aristotle
and the scholastics.

Specifically, it will be argued herein that an ontic predicate is a sim-

ple entity with a dual nature—one aspect a combinatorial state to or among
one or more subjects, the other aspect a content or intension (‘sense’) that
delimits as to kind and, when the predicate is polyadic, the number and or-
der of the unified subjects. The intension is also the source of a polyadic
predicate’s formal/logical properties (e.g., asymmetry, transitivity, reflex-
ivity), attributes absent in the limiting case of monadic properties. More
specifically, it will follow that 1) there is no ‘real distinction’, i.e., no plu-
ralizing composition, in a predicate between its combinatorial and inten-
sion aspects (in scholastic terms the distinction between the aspects is a
distinctio formalis a parte rei), and 2) the combinatorial state of a predicate
among a set of subject relata is unrepeatable, i.e., individuated. Under 1) a
predicate is a non-complex entity in the proper sense of a whole having no
third constituent in addition to the combinatorial and intension aspects that
unifies these aspects into the resultant predicate. Significantly and contrary
to tradition, conclusion 2) means that predication provides metaphysics with
a principium individuationis, an explanatory principle absent from popular
trope theory12 and from bundles-of-universals theories13, and one that is sus-

12
Campbell, Abstract Particulars; D. C. Williams, Review of Metaphysics 7 (1953): 3-18,
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

171-92; John Bacon, Universals and Property Instances: The Alphabet of Being (Oxford:
Blackwell, 1995); Peter Simons, ‘Particulars in Particular Clothing: Three Trope Theories
of Substance’, Philosophy and Phenomenological Research 54 (1994): 553-74, and ‘Fare-
well to Substance: A Differentiated Leave-Taking’, Ratio 11 (1998): 235-52.
13
Bertrand Russell, Inquiry into Meaning and Truth (London: Allen and Unwin, 19-40),
p. 98, and Human Knowledge, Its Scope and Limits (New York: Simon and Schuster,
1948), pp. 292-308. A. J. Ayer, ‘The Identity of Indiscernibles’ (1954), in Michael Loux,
Universals & Particulars, 2d. ed. (Notre Dame: Notre Dame University Press, 1976), pp.
263-70. Brand Blanshard, Reason and Analysis (LaSalle: Open Court, 1962), pp. 399-401.

tainable, unlike the ‘bare particulars’ of substratum theories14 or the indi-

viduation-by-instantiation of contemporary substance theories15. It follows
that the basic ontic units are individuated relation (including property) in-
stances Rni, each of which is a simple entity having the abstractable dual
aspects of outwardly directed unrepeatable predicability correlative with a
repeatable content or intension Rn. Intension Rn is numerically identical
across exactly resembling but distinct instances Rni, Rnj, Rnk, …. Due to its
nature as a combinator, a relation instance Rni exists only in the context of
a fact or state of affairs, :Rni(a1,a2,...,an), the latter being the unique com-
plex resulting from the unrepeatable predicable state of Rni among the dis-
tinct entities within the relata set,{a1,a2,...,an}. (The colon locution will be
used to distinguish a fact, if any, from a corresponding proposition.) Though
the specific relata set of a relation instance is the ‘occasion’ or secondary
cause of its individuation, it is its predicable union with these relata that is
the primary cause of its unrepeatability.

The combinatorial or predicable agency of relation instances, to-

gether with intension universals, are the potent features of this unit attrib-
ute ontology and what distinguish it from its chief rival, nominalistic trope
theory. Under trope theory individuated properties ‘free float’ in the sense
that they are by definition not predicable—each is a self-sufficing ‘little
substance’. Any gathering of tropes into a thick individual, e.g., apple a,
requires the posit of a specific relation to provide this unification, either
‘compresence’ (e.g., Campbell)16 or some internal ‘foundation relation’ in
the style of Husserl (e.g., Simons, Denkel)17. This predicably inert status is

The problem for bundles-of-universals theories is simple: a bundle of universals is itself

repeatable, i.e., universal, and so a bundle theory that admits only universals and bundles
thereof cannot account for individuation. Cf. Campbell, Abstract Particulars, p. 19.
14
For the bogus nature of ‘bare particulars’ see Campbell, Abstract Particulars, p. 7;
Loux, Metaphysics, pp. 113-17; Mertz, ‘Individuation and Instance Ontology’, Aus-
tralasian Journal of Philosophy 79 (2001): 45-61.
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

15
These theories do not give an account of individuation, they simply posit a category of
entities, substance universals, and assign to them the duty of individuation without expla-
nation how this is possible. How is it that a property numerically identical across a plural-
ity of individuals can, contrary to its repeatable nature, render these instantiating entities
unrepeatable? See Alica Rothstein and Nathan Oaklander, ‘Loux on Particulars: Bare and
Concrete’, The Modern Schoolman, 78 (2000), pp. 97-102.
16
Campbell, Abstract Particulars, pp. 21-22.
17
Simons, ‘Particulars in Particular Clothing’ and ‘Farewell to Substance’. See Edmund
Husserl, Logical Investigations (London: Routledge & Kegan Paul, 1970), p. 478. Arda Den-

plausible, though speciously so, for limiting monadic properties but not so
for polyadic relations that have an obvious existential dependence on, i.e.,
a ‘need to latch onto’, the members of relata n-tuples insofar as the rela-
tions function definitionally as relating. Hence the telling need within trope
theory of reducing relations to properties (Campbell) or treating relations
as, in effect, a radically different kind of entity from properties (Simons’
both non-predicable monadic as well as predicable polyadic tropes, and his
‘foundation’ relation which is not a trope at all). Both of these options are
errors and it is the implications of their correction that yield in great meas-
ure the instance ontology defended here.

The ontology of relation instances outlined above not only provides

solutions to the nexus of problems inherent in the Triple Aspect Problem,
but also provides a solution to a second and equally fundamental problem
for ontology. This is the Problem of Complexity:

What coherent account is there for the ubiquitous, fundamental epis-

temic given of organized wholes or complexes (structures, systems,
Gestalten) that are more than mere collections (‘heaps’, sets, mereo-
logical sums), i.e., plural wholes not necessitated by the mere exis-
tence of diverse entities (as are sets and mereological sums), but ex-
isting as a function of constituent inter-connections conditioned by
the specific natures or essences of the entities so structured?

Extending common experience, science tells us that reality consists of a di-

versity of hierarchical complexes: sub-atomic, molecular, biological, cog-
nitive, mathematical, etc. The common sense ‘continuous solid’ conception
of classical substances (humans, trees, statues, etc.) has been replaced by
one of wholes consisting of mostly empty space occupied by scattered enti-
ties, where the latters’ specific structuring is equally essential to the natures
of the wholes. These wholes have properties and enter into relations non-
existent at lower levels of complexity. It is intuitive and widely held that
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

what is needed to account for structural connections are polyadic relations,

but what is not appreciated is the explanatory power of this view refined to
include relation instances as the real and individuated connectors. Specifi-
cally, relation (including property) instances are necessary for and suffi-
cient as both ontology’s ‘primary substances’ and as the cause sine qua

kel, ‘On the Compresence of Tropes’, Philosophy and Phenomenological Research 57

(1997): 599-606.

non of all plural wholes whatsoever, and both notwithstanding instances’

ontic dependence upon their subject relata. For, contrary to the traditional
Aristotelian requirement on primary substances (Metaphysics 1028a20,
1028b34-1029a30), a view reiterated in contemporary ontology (e.g., in
Campbell, Armstrong)18, the existential dependence of instances on their
relata results not from some defect of being (ens), but rather derives from
their positive status as ontically productive and unifying principles. Indeed,
it is an argument below that atomic ontic entities must be predicable. This
and other results support the thesis that, as a single category, relation in-
stances of a diversity of intensions sustain a complete ontology charac-
terizable as network instance realism. Here from an atomic level of inter-
predicated instances hierarchical networks of cross-connected complexes
emerge that include at sufficiently structured levels ordinary ‘thick particu-
lars’, (e.g., apple a, Socrates, this chair) and more abstract complexes (e.g.,
minds, social structures, logical and mathematical systems). As complexes
emerge they themselves can be single relata for further emergent and sui
generis relation instances, the resultant networks being next-level com-
plexes. When the details are supplied for instance ontology, we would have,
I contend, an explanatorily adequate version of the thesis advanced by Witt-
genstein and recently argued by Armstrong that the world is a world of
facts, not things.19 Wittgenstein and Armstrong, however, hold that there are
no or few facts about facts.20 In complete contrast, the ontology for which
the following lays the foundation is what I take to be a Whiteheadian one
where the res verae are “individual facts of togetherness”21, i.e., relation in-
stances, and where the universe is a hierarchical system of systems com-
posed ultimately of inter-connecting and dependent relation instances,
atomic and emergent. Specifically, at the atomic level of this ontology there
exist only relation instances having other instances as relata—‘facts about
facts’—and higher ontic levels exist because certain relation instances have
complexes of facts as single relata—‘facts about systems of facts’.
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

18
Campbell, Abstract Particulars, pp. 98-99; Armstrong, States of Affairs, pp. 38, 99,
267.
19
Ludwig Wittgenstein, Tractatus Logico-Philosophicus (London: Routledge & Kegan
Paul, 1961), p. 7, Prop. 1.1; p. 13, Prop. 2.05. Armstrong, States of Affairs.
20
Wittgenstein, Tractatus, p. 13, Props. 2.061, 2.062; Armstrong, States of Affairs, pp.
139-147. For a critique of Armstrong’s independence thesis see my review of his States of
Affairs in The Modern Schoolman LXXV (1998): 227-31.
21
Alfred North Whitehead, Process and Reality (New York: The Free Press, 1978), p. 20,
also see pp. 27, 73.

2. The Problem of Ontic Predication and Locating the Predicable

Nexus

Ontic, or ‘material’ predication (hereafter simply ‘predication’) at the level

of fact is traditionally distinguished from linguistic/grammatical, or ‘for-
mal’ predication at the descriptive level of language,22 and refers to the na-
ture and cause of the nexus or union attributes (extra-linguistic or linguis-
tic) have with one or more entities they qualify. In contemporary terminol-
ogy, facts formed from material predication are the truthmakers for true af-
firmative propositions formed from formal predication. Providing a coher-
ent account of material predication is the fundamental Problem of Predica-
tion. Stated in its most general form the problem is:

22
It is instructive to note how nominalists who wish to eschew ontic predicates from
their ontology cannot retreat to nominalism’s ‘home field’ of language and there feel
secure. At the ‘formal’ level of grammar, Plato long ago observed (Sophist 262a-e)
that one could make no assertion with a string of only nouns (e.g., ‘lion stag horse’) or
only verbs (e.g., ‘walks runs sleeps’), but that one of each is required in order to ‘inter-
weave’ (e.g., ‘man understands’). The common correspondence view is that to assert
something as a fact requires the joining of a linguistic predicate to a linguistic subject,
and, when true, the assertion corresponds to the material predication of a property to a
subject entity or of a relation to multiple subject entities. For example, ‘Apple a is
round’, corresponds to a fact :Round(a) where property Round is predicably joined to
subject entity a. A nominalist might counter by saying that, despite the fact that lin-
guistic predication is pervasive and uneliminable in at least some languages, this is not
in itself evidence that it marks a fundamental aspect of reality. This objection misses
the mark. For, even at the level of grammar/syntax which is itself part of ‘reality’ there
are full syntactical relations—functioning as polyadic predicates—between terms (to-
kens if not types) from various linguistic categories, e.g., non-symmetric relations of
Linguistically-Predicable-of, Adjectivally-Modifiable-of, Adverbially-Modifiable-of.
Each such relation has a definite content that together with that of would-be subjects
determines whether or not they are predicably combined in a fact, as in :Linguistically-
Predicable-of(‘walks’,‘Theaetetus’), but there being no such fact corresponding to the
proposition Linguistically-Predicable-of(‘Socrates’,‘Theaetetus’), or in :Adverbially-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

Modifiable-of(‘beautifully’,‘sings’), but there being no such fact corresponding to the

proposition Adverbially-Modifiable-of(‘runs’,‘sings’). Even if these and like relations
are language specific, they are nonetheless relations, i.e., polyadic predicates. Rela-
tions are ubiquitous in the given of language (as they are in perception and cognition),
and hence like everyone else, nominalists must make use of polyadic predication. If, as
maintained here, it is true that ontic complexity is a reality (including multiple compo-
nents unified in the functioning of language), that relations are essential to complexity,
and that relations are not reducible to properties of their relata nor to relata themselves,
then relations, i.e., polyadic predicates, must be admitted into one’s ontology.

How are we to understand the predicational unification between an

n-place attribute with content or intension Rn and an n-tuple of sub-
jects, a1, a2, ..., an, that so unified constitutes a fact or state of affairs
:Rn(a1,a2,...,an), a whole where Rn mutually characterizes a1, a2, ..., an
and is thus distinct from and ‘more than’ the mere list Rn, a1, a2, ...,
an, or set {Rn,a1,a2,...,an}?

What, for example, is the difference between the fact that John loves Mary,
:Love(John,Mary), and the list John, Love, Mary, or the set {Love,John,
Mary}? What is an obvious relationship between the Problem of Predica-
tion and the Problem of Complexity is based upon the consequential fact
that facts or states of affairs are the simplest possible complexes. Under-
standing the unification constitutive of a fact is central to both problems.
As a preliminary, among the obvious differences between the unity of a
fact and that of a mere list or set are the following. 1) In a fact :Rn(a1,a2,
...,an) there is a non-arbitrary asymmetric order to the predicable nexus be-
tween subjects qua subjects a1, a2, ..., an, and the predicate qua predicate
(with intension) Rn, whereas the order in the list Rn, a1, a2, ..., an, is arbi-
trary and there is no order in the set {Rn,a1,a2,...,an}. This order is inde-
pendent of whether or not there exists an ordering among the subjects as
specified by the intension Rn. That there is a ‘direction’ to predication is
evident in the monadic case :P1(a) where the subject is a ‘thick’ particular,
e.g., apple a, for here a predicate, say Fruit, is predicable of a, i.e., it is the
case that a is a fruit, but not vice versa, i.e., it is non-sense to assert ‘Fruit
is an a.’ Importantly, the order in predication is evidence that the cause or
principle of the predicable nexus cannot be equally shared by subject(s)
and predicate, the nexus being ‘from one type to the other type’. When
predication is limited to the monadic it is not immediate where to attribute
the cause of the nexus, an ambivalence found in the tradition. However, if
polyadic predication is admitted full and unreduced, contrary to much of
Western ontology, then it is apparent that because a single fact can have
multiple subjects but not multiple predicates then predicates are the
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

cause/source of the predicable nexus. Appealing to a version of a classic

maxim [crucially, whose scope must be restricted—see Essays 2 and 4 in
this volume], unity is from the one, or unit, and its being ‘shared’ by the
thus unified. The single n-place predicate ‘shared’ among n relata is the
source of the unification requisite of the resulting fact. If, from the side of
the n subjects, each of the constituent subjects were singly the cause of the
factual unity with the single predicate then there would be n facts, not one.

And alternatively, if the n subjects were a unifying cause only collectively

then the unity of this collection would have to be accounted for by appeal-
ing to some further unifier distinct from each and all of the subjects as well
as from the predicate, and this leads either to a Bradley-type vicious re-
gress or to an incoherent ‘bare linking’. Both of the latter issues will be ex-
panded upon below. 2) For a contingent predicate, the predicable nexus,
and hence resultant fact, can come into and/or go out of existence, whereas
the corresponding set is atemporal ‘eternal’ in not being a relatum for tem-
poral relations. The predicable nexus in :Rn(a1,a2,...,an) is ontically distinct
from any of Rn, a1, a2, …, an, taken singly, or together in the manner of a
list or set. And 3), the existence of a fact whose predicate intension is Rn
precludes the existence of other facts with the same subject n-tuple but
whose predicate intensions are contraries of Rn, yet there is no such exclu-
sion among the corresponding lists or sets. What this says is that for the
predicable nexus the contents or intensions of subject(s) and predicate are
essential, whereas it is only their existence that is relevant to the corre-
sponding list or set. In regard to 3) a traditional assumption has been that in
any fact it is the predicate as an intension universal which is the cause of
the predicable nexus. This will be shown to be an error. Relevantly, it is to
be noted that in the above formulation of the Problem of Predication the
phrase ‘n-place attribute with content or intension Rn’ has been deliberately
chosen in lieu of what would be the more traditional ‘n-place universal Rn’,
i.e., where with the latter the nexus of attribution would be said to hold be-
tween the n-place universal Rn and an n-tuple of subjects. Methodologi-
cally, the adopted phrasing is motivated by its neutrality in regard to
whether predicable entities are universals or particulars. Further, however,
the now explicitly allowed predicate-as-particular option will emerge in the
following as cutting reality at its joints.

As a preliminary to an analysis of the predicable nexus proper, it is

necessary to examine proffered theories on where the cause of this unifica-
tion is located. Indicative of its obscurity in the tradition, philosophers
have variously located the principle of predicational unity: inside the fact,
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

outside the fact, and under the former as either a function of the predicate
or of the subject. Concerning the inside/outside debate, the dominant and
intuitive view has been that the predicable nexus is 1) an internal constitu-
ent of the resultant fact. However, in reaction to apparent problems with
this common sense view, recently philosophers have proposed that the
predicable nexus be assigned to either 2) the subsuming fact itself (e.g.,

Armstrong, Olson)23, or 3) to an agent external to the fact (recently by Val-

licella)24. Views 2) and 3) are erroneous, I shall argue, though instructively
so. In advocating 2) Armstrong is apparently motivated by his theses that a
fact does not exist when only its constituents exist and that the postulation
of a further unifying constituent leads to Bradley’s regress. In regard to the
first thesis, Armstrong holds that for a fact :P(a) its constituents are inten-
sion universal P and subject a, and, for contingent P, fact :P(a) might not
exist though P and a both exist. Vallicella rightly criticizes Armstrong’s
theory by observing that it is self-contradictory to require a fact to be both
a complex whole of parts and yet not a whole of parts insofar as it is more
than its constituents. The error can be seen in this way: for Armstrong we
have the three entities P, a, and :P(a), where fact :P(a) is to be the fact-
producing principle of unification among P and a, not the resultant of some
further unifier among them. Now, according to Armstrong, :P(a) = P + a +
(fact-producing principle of unification among P and a). Then by substitu-
tion, :P(a) = P + a + :P(a). Hence, fact :P(a) (on the left side) is identical to
something that is more than itself (the right side). Or alternately, fact :P(a)
is a proper part of itself, or, fact :P(a) is ontically prior to itself. Unfortu-
nately, however, Vallicella’s critique of Armstrong applies equally to his
own externalist thesis 3). To say that a requisite unifier is ‘external’ to its
fact :P(a) can only mean that it is not a constituent of its fact, and thus
:P(a) as the ‘sum’ of its parts has its being, whole and complete, independ-
ent of this unifier. But this is precisely what is denied in the required posit
of the unifier. The unifier essential to a fact :P(a) can be external to each of
the constituents construed here as intension P and subject a, but it cannot
be external to the subsuming fact of which it is of its defining essence—
fact :P(a) is definitionally P-and-a-factually-unified. In scholastic terms,
though the actuality or existence of an entity and its constituents may be
the effect of a sustaining external agent (e.g., God), what is of the essence
of an entity must be a constituent of it.

The error of both the subsuming-whole and externalist theories of

factual unity is even more obvious in the case of facts composed of non-
symmetric relations. Let R be a non-symmetric relation, e.g., Love, and let
it be the case that :R(a,b) and :R(b,a) both obtain. Each fact requires its
own unifier since either fact could obtain without the other (a fact whose
23
Armstrong, States of Affairs, p.118; Kenneth Olson, An Essay on Facts (Stanford: Cen-
ter for the Study of Language and Information, 1987), pp. 60-1.
24
William Vallicella, ‘Three Conceptions of States of Affairs’, Nous 34 (2000): 237-259.

expansion will be central below). But, if these unifiers are not constituents
of their respective facts, then, because the facts have otherwise exactly the
same constituents under Armstrong’s and Vallicella’s analysis, i.e., R, a,
and b, the facts must be identical, :R(a,b) = :R(b,a), which is absurd. This
conclusion is based upon assuming the Uniqueness by Composition Prin-
ciple:

Entities having exactly the same constituents are identical.25

Though this principle is intuitive and consequential (similar to the analo-

gous Axiom of Extensionality for set theory), Armstrong explicitly rejects
it,26 while Vallicella explicitly maintains both the principle and the thesis
that a fact is more than its constituents,27 an inconsistency on the above.
The difficulty here, and elsewhere as we are about to see, is the assumption
that in the subject(s)/predicate structure of a fact the predicate is an inten-
sion universal, and this independent of whether we assign the predicable
nexus to the subject(s) or the predicate. For, if this is assumed, then the
Uniqueness by Composition Principle does imply that :R(a,b) = :R(b,a),
for the same R, a, and b. Now it might be countered in the manner of Arm-
strong28 that universal R provides order among its relata by possessing dif-
ferent positions or ‘slots’ such that a occupies the first position and b the
second in :R(a,b), whereas a and b are reversed in positions occupied
within :R(b,a). The needed distinctness of facts is retained though Unique-
ness by Composition is violated. The trouble here is that if it is literally
and numerically the same universal R in both facts then its first position
would be occupied by both a (in fact :R(a,b)) and b (in fact :R(b,a)), and
mutatis mutandis for the second position. Even if one can make sense of
this, it implies that every such relation R, having a (and b) in both first and
second positions, is reflexive in regard to a (and b), which is counterfactual

25
See Mertz, ‘Individuation and Instance Ontology’.
26
D. M. Armstrong, ‘In Defense of Structural Universals’, Australasian Journal of Phi-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

losophy 64 (1986): 85-91; Universals: An Opinionated Introduction (Boulder: Westview

Press, 1989), pp. 90, 111; States of Affairs, p. 118-121. In the latter Armstrong allows for
a ‘weakened version’ of the Composition Principle that holds two complexes are identical
if and only if they have the same constituents and organization. The implication is that the
organization is distinct from the ‘constituents’. In contrast, the argument herein is that ‘or-
ganization’ is an aspect of a constituent of a fact, i.e., the predicate, and so the full
Uniqueness by Composition Principle holds.
27
Vallicella, ‘Three Conceptions of States of Affairs’, p. 248.
28
Armstrong, States of Affairs, pp.121-22.

(e.g., let R = Admires). Avoiding the false dilemma of having to give up

either the Uniqueness by Composition Principle (Armstrong) or the exis-
tence of facts (Lewis)29, the case will be made below that individuated rela-
tion instances are the true predicates in facts and that intensions are shared
aspects of the instances. It will be seen that the equivocation of non-pred-
icable intensions with their subsuming predicable instances is what makes
Bradley’s regress plausible.

The internal placement of the cause of predicational unification must

be retained, i.e., the principle of unification of a resultant fact must be a
constituent of the fact. Doubts to the contrary have been a function of the
problems that arise in attempts to assign predicability to constituents of a
fact as traditionally conceived. For a monadic fact :P(a), which has been
the traditional paradigm (a distorting limitation on the following), it has
been one part of the tradition to assign predicability, directly or indirectly,
to the intension P. The weakness here is not a function of a commitment to
intension universals. Indeed, I propose that there are compelling arguments
for universals,30 and offer the following as indirect evidence for them.
Rather, the deficiency of the traditional analysis of facts arises from its
commitment to either a) that universals are themselves predicable entities
(the traditional ‘Aristotelian’ view, the widely approved Fregean theory of
‘concepts’ being a modern explication)31, or b) that universals are them-
selves not predicable but rather work mutually with either an exemplifica-
tion (or instantiation) relation (e.g., Moreland)32 or with ‘non-relational
ties’, i.e., intensionless linkings (e.g., Bergmann, Strawson, Hochberg)33, to

29
David Lewis, ‘Comments on Armstrong and Forrest’, Australasian Journal of Philoso-
phy 64 (1986): 92-93. Also see his Parts of Classes (Oxford: Blackwell, 1991), pp. 56-57.
30
E.g., as found in Armstrong, Nominalism and Realism, and Reinhardt Grossmann, The
Categorical Structure of the World (Bloomington: Indiana University Press, 1983).
31
Gottlob Frege, Translations from the Philosophical Writings of Gottlob Frege, ed. P.
Geach and M. Black (Oxford: Blackwell, 1970), pp. 46-47, 54-55, 152. Also see Frege,
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

‘On the Foundations of Geometry’ in Essays on Frege, ed. E. Klemke (Chicago: Univer-
sity of Chicago Press, 1968), pp. 569-71.
32
Moreland, Universals, Qualities, and Quality-Instances, pp. 170, 194. Also see his ‘Is-
sues and Options in Exemplification’, American Philosophical Quarterly 33 (1996): 133-
47.
33
Gustav Bergmann, Realism (Madison: University of Wisconsin Press, 1967), pp. 9,
42ff. Herbert Hochberg, ‘A Refutation of Moderate Nominalism’, Australasian Journal of
Philosophy 66 (1988): 188-207. P. F. Strawson, Individuals (London: Methuen, 1971), pp.
168ff.

form wholes predicable of n-tuples of subjects. There are severe problems

specific to each of the alternatives, a) and b). First, against alternative b) is
the fact that the hypothesizing of a non-relational tie is equivalent to posit-
ing a ‘bare linking’, analogous to and as bogus as a ‘bare particular’.34 This
point will be developed below in the context of a further demonstration of
the untenability of alternative a)—that universals themselves are the cause
of their predicable union with their subjects. There it will be argued that
predicates qua predicable are necessarily unrepeatable. A further and rele-
vant critique against alternative b) is the observation that there is nothing
unique to an especially appointed exemplification relation in its unification
of its relata that is not possessed by any other relation R insofar as R is
predicable of, i.e., ‘actually relates’, its relata. This is a lesson of Bradley’s
Regress, as reviewed below, and it applies equally to the limiting case of
monadic predicates. In general the question becomes: Is it reasonable that,
if upon a certain ontological analysis (e.g., the construal under Bradley’s
regress of predicates as non-combinatorial; or the Wiener-Kuratowski ex-
tensionalist reduction of relations) it is concluded that entities of a category
(e.g., properties and relations) lack a specific characteristic (e.g., a linking
state among their subjects; or an intensional content), to then posit an en-
tity from the same category (e.g., an exemplification or instantiation rela-
tion; or the Element-of relation) as exempt from the analysis and as having
the explanatorily needed characteristic? Clearly, the answer is no, and the
fact that there is pressure to posit such an entity as exempt is evidence that
the ontological analysis of the subsuming category is erroneous. Conse-
quently here, either all ontic predicates, properties and relations, are com-
binatorial among their relata or none are, including exemplification. (The
direct relevance of intensions will be seen below.) This critique applies
equally to contemporary trope theory where, as a surrogate for monadic
predication, non-trope unifying relations—Compresence or ‘founding’ re-
lations—are posited to account for the ‘bundling’ into thick particulars of
non-predicable tropes, and, as a surrogate for universals, the non-trope Re-
semblance relation is posited as what links tropes in being intensionally
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

‘the same’. Yet, if these dyadic relations (or their instances) are predicable
then why not monadic tropes, e.g., Redi, Roundj, Has-massk? To avoid this
problem, Campbell in advocating trope theory is consistent in having ar-
gued that polyadic relations are reducible to monadic tropes.35 But then un-
der such a reduction to non-predicable tropes the predicable/intension-
34
See note 14.
35
Campbell, Abstract Particulars, pp. 97ff.

controlled-linking aspects of the Compresence/founding and Resemblance

relations are eliminated, and yet it is precisely these aspects that give these
relations their efficacy for the theory and is, if implicit, the raison d´être
for their posit. That the polyadic is not reducible to the monadic will be
taken up latter.

There is a more consequential critique of the present assay of singu-

lar facts into subject individuals and predicate universals, one independent
of whether the predicable nexus is assigned to the subject or the predicate,
or whether the predication is immediate as in a) or mediate as in b). Suc-
cinctly, the assay requires that an intensioned or ‘thick’ individual a reduce
upon analysis to an incoherent qualityless substrate—a bare particular. The
argument is straightforward: If every characteristic, quality, or attribute
whatsoever of an individual is a predicate of it and on the present assump-
tion a repeatable universal, then an individual as unrepeatable must be dis-
tinct from all its characteristics—it must be a bare particular. This is appar-
ently Aristotle’s conclusion at Metaphysics 1029a10-30 where he reduces
to absurdity his ultimate-subject criterion for primary substance given in
the Categories (1b10, 2a34),36 i.e., the absurdity of the thesis that primary
substances are non-predicable individuals, and every other kind of entity is
predicable of a primary substance. The conception of primary substance
here is analogous to that of the naive ‘vessel’ model of subjects (cf. Catego-
ries 15b17ff; Metaphysics 1023a6ff) where as a ‘container’ the substance
subject serves to unify by gathering within it all its attributable characteris-
tics. The analogy is one of a sack as it holds together its contents. Another
analogy satirized by Russell is thinking of a thick particular as something
like an individuating and unifying peg on which is hung all the particular’s
characterizing universals. Yet, whether considered analogous to a container
or to a focal unifier, an entity distinct from absolutely all its attributes re-
duces upon analysis to a bare particular, which is, euphemism aside, to
evaporate into non-being. For the mis-analogy of the sack or the peg is that
either would have attributes which on the present analysis must be re-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

assigned from the sack to the content of the sack or from the peg to the uni-
versals hung on the peg—sack and peg reduce to no-thing.

Notwithstanding contemporary attempts to salvage the doctrine by

positing an external ‘tied-to’ relation between bare particulars and their es-
36
See Mary Gill, Aristotle on Substance (Princeton: Princeton University Press, 1989), pp.
19ff.

sential properties (e.g., Moreland)37, bare particulars remain in themselves

devoid of all characteristics whatsoever, and this is their incoherence. The
theorized ‘tied-to’ relation itself indicates how for bare particulars even
their ‘essential’ attributes (e.g., simplicity, unrepeatability) are radically
external to them, displaying how bare particulars can have absolutely no
content or intension. Hence, even their raison d’être—their unrepeatabil-
ity—is outside their constituting essence, and hence, incoherently, some-
thing they need not be. This and other arguments against the coherence of
such a concept are, I contend, compelling (e.g., those of Campbell, Loux,
and Mertz)38, and I refer the reader to them. Of these additional arguments
I would here mention two. First, based upon the above Uniqueness by
Composition Principle that entities having exactly the same constituents
are identical, there could be at most one bare particular, just as analogously
there can be but one null set. Entities having absolutely no constituents
[proper or improper (in having no essences)] have by default the same con-
stituents and are thus identical. It follows, absurdly, that there could be but
one ordinary thick individual.* Secondly, the predication of properties of a
subject is a function of the compatibility of the contents or intensions of
each of the properties and that of the subject. Even so-called ‘external’ re-
lations presuppose intensions characterizing their relata, e.g., the external
spatial relation in the fact :Left-of(a,b) requires a and b have (or have con-
tent that support) the content Being-Extended, it not being possible for the
relation Left-of to take as subjects any entities whatsoever, e.g., a and b
could not be the numbers 3 and 4. Now, a would-be subject devoid of all
intension or content could not condition or delimit properties as to their
predicability or nonpredicability of it. Hence, the predicates of a bare par-
ticular would be every one arbitrary and contingent, contrary to its neces-
sary properties such as simplicity and unrepeatability.

The important lessons to be learned from the above is that a slide in-
to a reductio of bare particulars results from the assumptions that a predi-
cate (immediate or mediate) is an intension universal, and that any content-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

or nature-bestowing intension possessed by an individual is predicable of

it. As universals, every characteristic of an individual is other than it ren-

*
[Moreland and Timothy Pickavance have responded to this argument in ‘Bare Particulars
and Individuation: Reply to Mertz’, Australasian Journal of Philosophy 81 (2003): 1-13.
My counter-response is found in the same issue.]
37
Moreland, ‘Theories of Individuation: A Reconsideration of Bare Particulars’.
38
See note 14.

dering it a bare particular. These assumptions require of an ontology that

every individual that is not a bare particular be a complex entity, and force
bare particulars to be constituents of whatever is the ontology’s lowest
level intensioned particulars (e.g., substance in the Categories; Arm-
strong’s states of affairs; Moreland’s unit attributes). Consequently, to
avoid these difficulties an ontology is needed where what is predicable is
particular, and where atomic individuals have as their qualitative contents
repeatable intensions that are not predicable of them, but where the unity of
the individuating and intension aspects requires no additional linking predi-
cates constitutive of the atomic individuals. (To reiterate, this latter re-
quirement is what is meant by a simple entity—not that distinct and par-
tially characterizing aspects cannot be abstracted from the entity, but that
these aspects do not exist in it as subjects of a further predicable unifier
that is also part of the entity.) The meeting of each of these requirements
will be seen to follow from the defining combinatorial character of predi-
cates. It is to be noted that Aristotle in the Metaphysics went some distance
in satisfying these conditions with his doctrine of unit substantial forms.
Here a substantial form is individuated (Metaphysics 1042a29, 1049a35,
1071a25-30), it is predicable of multiple secondary matter and in this is the
cause of a resultant structured whole—hence the ontic equivalent of a
polyadic relation (Metaphysics 1016b12-17, 1041b1-30, 1045a6-25), and
there is no ‘ontic distance’ between its combinatorial mode and its inten-
sion (implicit in Aristotle (cf. Metaphysics 1045b8-24), but explicit in the
scholastics, e.g., as referenced in Suarez39). Unfortunately, the explanatory
virtues of this analysis are negated in Aristotelian ontology by the individu-
ating role assigned to matter, which in the limit of prime matter must be-
come a bare particular, and by the official reduction of all polyadic to mo-
nadic predicates as reinforced by a false containment model of predication.

3. The Erroneous Containment Model of Predication

It has been argued that there cannot exist bare particulars, i.e., every unre-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

peatable individual must possess in a manner more intimate than external

predication a content of one or more repeatable intensions. To make inten-
sion universals characterizing an individual externally predicable of and
hence numerically distinct from it is to introduce a separation fatal to the
39
Francisco Suarez, On the Various Kinds of Distinctions (Disputationes Metaphysicae,
Disputatio VII, de variis distinctionum generibus), trans. Cyril Vollert, S.J. (Milwaukee:
Marquette University Press, 1947), I, 17-29, pp. 28-38.

individual qua individual. The alternative is to eliminate the ‘ontic dis-

tance’ between an individual and its intensions by placing the latter some-
how ‘in’ the individual. We shall see that this is indeed possible and re-
quired for unit attributes with their single intensions, as based upon results
below that the combinatorial and intension aspects of a predicate are ab-
stractions from a simple entity and that the combinatorial aspect individu-
ates the whole. However, as we shall see below this requirement of col-
lapsing the ontic distance between an individual and its intensions is in-
compatible with the assumption that the individual is characterized by mul-
tiple intensions. Seemingly in line with the leading thesis is the long-pre-
vailing but in fact insidious containment model of predication where a
predicate is held to be ‘contained in’, ‘immanent in’, ‘inherent in’, its sub-
ject. It is held, for example, that apple a has mass because the property
Has-Mass is ‘in’ a, a plausible theory given that apple a is a spatially ex-
tended object. The theory is less plausible, however, for non-spatial and
apparently non-complex subjects, e.g., that the property Prime is ‘in’ the
number 3.40 The containment model was explicit in the semantics of the
medieval inherence theory of the copula. In a standard categorical proposi-
tion such as ‘Apple a is round’ the copula ‘is’ was considered to be a sig-
num inhaerentiae of the universal connoted by the grammatical predicate
‘round’, it having the status of inesse subjecto, here the containing subject
being apple a.41 Historically, the containment model has remained mostly
implicit with its further implications either unrecognized as such or unat-
tended. Two notable exceptions are the implications Leibniz drew from his
explicit Praedicatum-inest-subjecto doctrine42, and those drawn by Bradley
from his thesis that every judgment is a differentiation of the predicate out
of a subject which is then ‘restored’ to the subject by predication, the sub-
ject of every judgment being ultimately all of reality or the Absolute43.

40
The number 3 is at most a simple intension, or at least, under structuralist theory, an
empty place holder in an ordered structure that is the natural numbers system, and in
either case no amount of attempted conceptual dissection could find properties Prime,
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

Odd, etc., a literal interpretation of Kantian ‘analyticity’ notwithstanding.

41
See John Malcolm, ‘A Reconsideration of the Identity and Inherence Theories of the
Copula’, Journal of the History of Philosophy 17 (1979): 383-400. For a particular ex-
ample of the inherence theory in Aquinas see Keith Buersmeyer, ‘Predication and Par-
ticipation’, The New Scholasticism 55 (1981): 35-51.
42
Gottfried W. Leibniz, Philosophical Papers and Letters, 2d. ed., ed. L. Loemker (Dordrecht:
Reidel, 1969), pp. 264, 334, 337.
43
F. H Bradley, Appearance and Reality, 2d ed. (1897: reprt. ed., Oxford: Clarendon Press,
1966), pp. 144-45, 147-49.

Under the containment model a multiply characterized subject can be

conceived as either 1) a plural heterogeneous bundle, e.g., a bundle of
tropes (Campbell, Simons, Denkel, LaBossiere)44, or as 2) a monistic ho-
mogeneous ‘blend’ of its properties, e.g., analogous to gray paint as the
identity-losing blend of black and white paint. The notion of a constituent
blend is obscure, though I take it to be what is ultimately the implied nature
of Bradley’s Absolute45 and as required by any monism. I shall attempt
briefly to characterize it. According to Bradley, “Since diversities exist…
they must be united. …But this means that A and B are united, each from it
own nature, in a whole which is the nature of both alike. And hence it fol-
lows that in the end there is nothing real but a whole of this kind.”46 In
‘blending’ each element is ‘transformed’ effecting a ‘transcendent’ whole,
“A whole in which distinctions can be made [by abstraction], but in which
divisions do not exist.”47(my insert) The entities that are blended each lose
their identity by a kind of mutual suffusion into a resulting whole, and this
in such a manner that when these entities are intensions they are no longer
truly assertable of the whole; they are “characters of a Reality which they
cannot express”48, analogous to the fact that blended gray paint is literally
neither black nor white. In effect, under the blending version of contain-
ment we have unity with no real plurality.

Now, inherent to the containment model, predication as a state of un-

ion between a subject and its characterizing attributes ceases to be ontically
primary, but is rather now secondary and derivative of what must be a
more fundamental unification existing among the attributes and whatever
else constitutes the subject (e.g., an individuator if the other constituents
are universals). Yet, here neither the blending or bundling forms of unifica-
tion are sustainable. Specifically, in opposite ways blending and bundling
each contradict the facts that both diversity and unity among the diverse is
real (i.e., complexes are real), and that there are irreducible polyadic rela-
tions that obtain among the diverse. In regard first to Bradlarian blending
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

we have seen that it implies the counterfactual thesis that diversity is an il-

44
See notes 16 and 17. Michael LaBossiere, ‘Substances and Substrata’, Australasian
Journal of Philosophy 72, 3 (1994): 360-70.
45
Bradley, Appearance and Reality, pp. 154, 404, 509-10.
46
Ibid., p. 510.
47
Ibid., p. 128. Also see pp. 161, 519, 521.
48
Ibid., p. 511.

lusion. Relatedly, since asymmetric relations imply an order and hence a

distinction between relata, they too must be declared illusion. Let it be the
case that :R(a,b) exists for asymmetric relation R. As a ‘predicate’ of both
a and b, R must blend into both and so carry with it the mutual blending of
each relatum with the other. As Bradley puts it, “A relation must at both
ends affect, and pass into, the being of its terms”49 and so “A relation is un-
meaning, unless both itself and the related are the adjectives of a whole.”50
Yet this forces a loss of distinction between relata a and b that the asym-
metry of R presupposes. In Bradley’s words, “The terms and the relation
must ‘enter’ one into the other, and yet again are ruined if they do so.”51
Bradley interprets correctly the consequences of blending as implying the
absurdity of both polyadic and monadic predicates, blending being for
Bradley the only way to account for predication and its requirement of a
real distinction between subject(s) and predicate. In total, “All predication,
no matter what, is in the end untrue and in the end unreal…”52 Hence, all
plurality is an illusion, and no proposition is in the end true. In effect, Real-
ity, the ‘One’, becomes the cohesive at the loss of the cohered. Here unity
is from the one insofar as it obliterates all distinctions among the united.
But then of what is this a reductio? Surely if something is to be rejected, it
is to be the mysteries of blending and monism over the pervasive given of
the diverse and related.

Under the alternative bundle theory where real diversity is a given,

the standard and necessary account of the needed unity is the posit of a
special relation (e.g., Compresence, ‘founding’) unifying the constituent
properties F, G, H, … into resultant a (the constituents including necessar-
ily at least one individual, e.g., a trope, if resultant a is an individual). But
then, the unity here is supplied by a relation in its role as a polyadic predi-
cate, and under the containment model of predication all multi-subject
polyadic predicates must reduce necessarily to single-subject monadic
properties of their relata. For, containment taken literally requires that the
predicable nexus a property (or property instance) P has with an ‘encom-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

passing’ subject a be independent of and not conditioned by P’s simultane-

ous predicable nexus with any other subject b, b ≠ a. That is, consistent

49
Ibid., p. 322. Also see p. 347.
50
Ibid., p. 394.
51
F. H. Bradley, ‘Relations’, in Collected Essays, Vol. 2 (Westport: Greenwood Press, 1970),
p. 638.
52
Ibid., p. 672.

with the containment analogy, property P’s being wholly ‘in’ a implies that
no part of P-insofar-as-it-is-predicated-of-a is ‘left outside of’ a to be
shared by, and in this be an ontic bridge to, some other subject b. Conse-
quently, the property reduction of relations implied by a strict interpreta-
tion of the containment model of predication itself implies that the would-
be bundling relation between, say, constituent properties F and G must re-
duce without remainder to foundational properties ‘in’ each, and hence the
requisite cross-constituent unification essential to resultant a would disap-
pear into the then mutually isolated F and G. No real interposing connect-
edness would result. This strict implication is counterfactual.

The latter implication is no doubt why historically a modified reduc-

tion of relations inspired by the containment model became the standard
analysis. Here a real relation (held to be at most dyadic) was said to be re-
ducible to monadic properties of each of its relata, the relation’s esse in,
but where the latter each have a sui generis character of ‘being toward’ the
other paired relatum, the relation’s esse ad.53 The ‘toward’ or esse ad as-
pect was admitted as a minimal concession to our intuition of and theoreti-
cal need for inter-subjects connectedness. Though it was historically endur-
ing, the esse in/esse ad analysis is untenable, as signaled by the counter-
intuitive theories that have been drawn from it. For example, in his rigor-
ous championing of the praedicatum-inest-subjecto tradition, Leibniz ridi-
culed the notion of an unreduced relation as an attribute having to have
“one leg in one [relatum] and the other [leg] in the other [relatum]”54(my
inserts), and, pressing the doctrine’s implications, concluded to an ontol-
ogy of isolated monads, each internally ‘mirroring’ the rest of the world.
Since every entity enters into at least some relations with every other, each
of Leibniz’s monads would contain the esse ad’s (the ‘reflections’) for
every one of these relations without being actually linked to any other en-
tity. Importantly and the reason for its failure, the ‘being toward’ aspects of
this reduction would have to be themselves intensionless, for otherwise
they would be further full relations calling for further esse in/esse ad re-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

ductions, and so on to vicious regress. A reductio of these ‘bare linkings’

will be given below.

53
Mark Henninger, Relations: Medieval Theories, 1250-1325 (Oxford: Clarendon Press,
1989).
54
Leibniz, Philosophical Papers and Letters, p. 704.

Within this tradition it was but a short step to relegating the distinc-
tive esse ad aspects of the reduction, and consequently relations generally,
to the status of products of the mind (entia rationis). The ‘toward’ aspects
became for some philosophers (e.g., Hume)55 free or blank associations re-
sulting from ‘comparisons’ of the mind but revealing no ‘real connection’,
i.e., no unity controlled by the compatibility of relata and relation inten-
sions, the latter having been abstracted away into the esse in’s. Among the
notable casualties of this tack is a loss of the nomic necessity of causation,
it being replaced with the residual alternative of contingent constant con-
junction. Here we have plurality with no real extra-conceptual modes of
connectedness, and no connections conditioned on anything more than the
existence of their terms and the unifying agency of the associating mind. In
an attempt to eliminate the subjective/ideal character of unity by cognitive
agency, an extensionalist/nominalist tack is to adopt the strengthened thesis
that the existences of entities, independent of their natures, is sufficient
tout court for the existence of resultant wholes of all possible combinations
thereof. Here either set theory or mereology is awarded a primary onto-
logical standing, and ontology proper reduces to models, i.e., substructures,
from one of these formal theories. Yet, these formal models are themselves
‘extrinsically determined’ substructures constructively selected, i.e., differ-
entiated out, by the agency of minds. What would be in fact intensionally
determined relationships making up a given a posteriori complex, some
contingent, some necessary, guide the unofficial but prerequisite creative
act of selecting out the primitive sets or sums of the a priori model. The
model itself consists of the structure containing only modally necessary re-
lations following upon these atomic sets or sums and that are definable in
the terms of either the Element-of relation for set theory or the Part-of rela-
tion for mereology. The essential role of mind is not eliminated, just offi-
cially ignored. Equally telling is the fact that here the role of intensions has
not and, indeed, cannot be eliminated from the theories. For, in either the-
ory the primitive relation, Element-of or Part-of, is essentially intensional
since the axioms of the theory are not random, i.e., their negations are not
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

alternatively assertable, but rather are necessitated by the very intension or

content (the ‘meaning’) of the relation, and the relation is in principle un-
reducible to any kind of extensional surrogate within the theory. Conse-
quently, the extensionalist strategy of identifying complexes with formal

55
David Hume, A Treatise on Human Nature, I. ii. 4; ed. T. H. Green and T. H. Grose
(1886; rpt. ed., Germany: Scientia Verlag Aalen, 1964), pp. 352, 559.

models avoids neither the necessity of an appeal to cognitive agency nor

the necessity of connectedness defined and delimited by intensions.

In sum, neither the blending nor bundling alternatives of the con-

tainment model of predication can provide an account of how a thick indi-
vidual can be characterized by multiple attributes. In opposite ways each
alternative proves contrary to the fact of real unity among the truly diverse.

4. Relations and the Combinatorial Nature of Predication

Historically, the property reduction of relations was an error whose correc-

tion was slow in coming. In the formal sciences of logic and mathematics
the explanatory value of relations as full multi-subject predicates was ex-
plicit from at least the middle of the nineteenth-century.56 However, it was
not until Russell’s Principles of Mathematics, 1903, that we have what is
now the locus classicus for arguments showing the ontological impossibil-
ity of the monadic reduction of polyadic predicates.57 These arguments turn
on the order of the predication that some relations (e.g., asymmetric and
non-symmetric) maintain among their relata, in each case an order that
cannot be eliminated extensionally (cf. Hochberg)58 but is rather a function
of the relation’s intension.59 Recently I have reinforced Russell’s argu-
ments against counter-claims found in Campbell’s version of contemporary
trope theory.60 I shall not rehearse these arguments here. The crucial point
for ontology is: Relations are unreducible to properties. Relations are not
characterized as esse in/esse ad, but rather as esse inter—as fully real ‘in-
terconnectives’ among their relata.

56
There were explicit attempts in the sixteenth and seventeenth centuries to demon-
strate that proofs in Euclid’s Elements are reducible to syllogistic form. See Neal Gil-
bert, Renaissance Concepts of Method (New York: Columbia University Press, 1960),
pp. 89-90. Ian Mueller has shown that not even the first proposition of the Elements is
reducible to syllogistic form. Ian Mueller, ‘Greek Mathematics and Greek Logic’, in
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

Ancient Logic and Its Modern Interpretation, ed. John Corcoran (Dordrecht: Reidel,
1974), pp. 35-70.
57
Bertrand Russell, The Principles of Mathematics, 2d ed. (1903: reprt. ed., New York:
Norton, 1938), pp. 221ff.
58
Herbert Hochberg, ‘The Wiener-Kuratowski Procedure and the Analysis of Order’,
Analysis 41 (1981): 161-3.
59
Bertrand Russell, My Philosophical Development (London: Allen & Unwin, 1959),
p. 67.
60
Mertz, Moderate Realism, pp. 163-71.

It is important to note that the non-reducibility, non-eliminability of

ordering relations is strong evidence against the would-be deflationary the-
sis that necessary or ‘internal’ relations are supervenient upon and so rep-
resent no ontic addition over and above their relata (e.g., Campbell, Arm-
strong)61. For, given, say, the numbers 2 and 3 it is necessarily the case that
the Less-than relation exists between them, i.e., necessarily :<(2,3), yet on
the arguments cited the Less-than relation is not reducible to monadic
properties of 2 and 3 or anything else. In particular, the ordering of the
Less-than relation cannot be accounted for by any monadic property, or set
or mereological sum thereof. The only seeming alternative to admitting an
uneliminable entity into one’s ontology is an argument showing that the
very concept of such an entity is incoherent and is therefore illusion or
‘appearance’ in the relevant sense of Bradley’s arguments against relations
generally. Yet authors who advance the supervenient view do not want to
grant such ‘unreality’ to internal relations.

Relative to traditional substance/attribute ontology, an analysis of the

nature of relations motivated by their non-reduction produces a kind of
‘Copernican Revolution’ in both analytic power and ontological gain. In
broad terms what is gained is the liberating perspective where, as it were,
the monadic tail no longer wags the polyadic dog. Monadic properties are
seen to be but the limiting case of polyadic relations, where the latter have
ontically and logically significant characteristics obscured or absent in the
former (e.g., ordering and formal properties such as asymmetry).62 Specifi-
61
Campbell, Abstract Particulars, pp. 97ff. Also see Campbell’s ‘The Place of Rela-
tions in a Trope Philosophy’ in Proceedings of the Colloque International de Philo-
sophie de Grenoble: La Structure du monde; Objets, Propriétés, Etats de choses, in
Recherches sur la philosophie et le langage, Universite Pierre Mendes France, Grenoble,
2003; Armstrong, States of Affairs, pp. 87-88, 92-93.
62
Also globally, the non-eliminability of relations provides the ontic underpinning for
the analytic methodology that holds that to discover the relations an entity a has with
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

other interrelated entities, i.e., to find a’s ‘position’ in the context of a complex, is to
have an explanation of a. As a slogan, ‘To explicate is to interrelate.’ This contrasts
with the once common methodology that ‘To explicate is to eliminate.’ See Peter
Strawson, Subject and Predicate in Logic and Grammar (London: Methuen, 1974), p.
37, 54-55. Also see Jerrold Katz, The Metaphysics of Meaning (Cambridge: MIT Press,
1990), p. 186. The containment model of predication, requiring as it does the property re-
duction of relations, not only abets the latter and, I claim, impoverished reductive meth-
odology, but, insidiously, eliminates relations as the corrective to it. See Moderate Real-
ism, p. 37.

cally, the non-reduction of relations to properties motivates, first and pri-

marily, the recognition of what is the sui generis unifying nature of predi-
cates among their subjects—predicates are combinators, i.e., agents of uni-
fication. This point will be established presently. The combinatorial nature
of predication is what Frege attempted to capture with his metaphorical as-
cription to concepts of an ‘unsaturated’ character. As unsaturated, a con-
cept is “capable of serving as a link”, for in every complex whole “at least
one [constituent] must be ‘unsaturated’ or predicative; otherwise they
would not hold together.”63(my insert) Medieval philosophers noted the
necessity of the combinatorial aspects of forms and properties in order for
them to be “per se and directly joined without the medium of another
mode”, here referred to as ‘modes’ (Suarez)64 and by others as ‘disposi-
tions’ (Buridan)65. Recently, Armstrong has adopted the ‘way’ metaphor
suggested by David Seargent where, in response to Bradley’s Regress, it is
said that properties and relations as ways “form so much closer a unity
with the things involved that perhaps further relations are not required.”66

The unifying nature of polyadic relations is ontically primitive, and

necessarily because of this is accessible only by a kind of conceptual paral-
lax focusing attention on and differentiating out this fundamental aspect of
reality. The argument now known as Bradley’s Regress is a classic and
perennial heuristic device cited as effecting the insight that relations have a
combinatorial nature.67 Also instructive in making this same point, I pro-
pose, is this argument’s interpretation, e.g., by Bradley and Ockham68, as
showing the self-contradictory nature of relations. Central to the analyses
herein is understanding the grounds for these two interpretations. Bradley’s
Regress argument is as follows. In a fact :P1(a), e.g., Plato is bald, or for-
mally :Bald1(Plato), either property P1 is predicationally attached to subject

63
Frege, Translations, pp. 54-55.
64
Francisco Suarez, On the Various Kinds of Distinctions, I. 20, p. 31.
65
John Buridan, Questiones in metaphysicam aristoteles, V, Q. 8, fols. 31, 33. Por-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

tions translated by Calvin Normore, ‘Buridan’s Ontology’, in How Things Are, ed., J.
Bogen and J. McGuire (Dordrecht: Reidel, 1985), p. 198.
66
Armstrong, States of Affairs, pp. 30, 38, 98.
67
Cited as displaying the unifying aspect of at least monadic properties, the regress
argument is anticipated in Aristotle (in regard to substantial form), and found fully in
philosophers such as Avicenna, Buridan, Suarez, Frege, Russell, and most modern on-
tologists. For references see my Moderate Realism.
68
Bradley, Collected Essays, pp. 628-76. William of Ockham, Ockham’s Theory of Terms,
trans. Michael Loux (Notre Dame: University of Notre Dame Press, 1974), p. 170.

a or it is not. If not, then the unity between P1 and a is not that of attribu-
tion but rather at best mere association, and in which case fact :P1(a) re-
duces to the juxtaposition of set {P1,a}, e.g., {Bald1,Plato}. Yet, clearly
fact and set are not identical—in the example set {Bald1,Plato} no claim is
made as to whether Plato is bald or not, yet with the complex :Bald1(Plato)
we have the fact grounding the affirmative claim. Moreover and as previ-
ously noted, for contingent P1 the fact :P1(a) can come into and go out of
existence whereas the set {P1,a} exists necessarily and so is ‘eternal’. In-
deed, the set exists even when for some intensions (e.g., Unicorn, Phlogis-
ton) no parallel fact ever obtains. In regard, then, to the initial fact :P1(a),
an additional unifier, say relation R2, is required to supply the predica-
tional unity needed for the original fact, the latter now becoming :R2(P1,a),
e.g., :Exp2(Bald1,Plato), where ‘Exp’ designates Exemplification. Now the
same question can be asked of this new fact: Is R2 predicationally linked
with P1 and a? If not, then fact :R2(P1,a) becomes set {R2,P1,a}, which re-
quires a further relation R3 to supply the requisite predicational unity, i.e.,
the original fact has now become :R3(R2,P1,a), e.g., :Exp3(Exp2,Bald1,
Plato). Clearly this denial of a predicational unification between predicates
and subject(s) leads to vicious regress. And, to admit this predicational
unity at any step implies its existence at the first step, :P1(a), the regress
thus eliminated ab initio. Points to be observed here are that, first, at each
step where predicational unity is required, it is intuitive that a polyadic re-
lation should be called upon to do the job, and rightly so since the alterna-
tive of intensionless or bare linkings is incoherent as will be seen. Further
and importantly, the initial plausibility of the ratcheting steps of the regress
imply that we intuitively distinguish two ‘versions’ of an n-adic predicate:
one as predicable in a fact, i.e., as a ‘linking’ among n relata under an in-
tension, e.g., a-loving-b, and the other as the intension abstracted from its
predicable aspect, e.g., Love. The regress plays on this possibility, and at
each n+1-step it speciously equates the abstracted ‘predicably inert’ inten-
sion with what was in the n-step the full predicate as intension-plus-nexus-
among-relata. That is, at each n+1-step the regress takes away the conjunc-
tive or unifying aspect of the then relation, an aspect essential to the n-step,
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

and leaves as a residue the relation intension as an unattached ‘between’—

as but one more element to be unified. Bradley is explicit in this: “An ac-
tual relation… must possess at once both the characters of a ‘together’ and
a ‘between’, and, failing either of these, is a relation no longer…. On the
other hand, if, to remain themselves, our terms retain their character as in-
dividuals, there is no legitimate way (we have seen) to their union in fact.
We are without the ‘together’, which (like the ‘between’) is essential if any

relation is to be actually there.”69 Hence, the straightforward interpretation

of what the regress demonstrates is that we cannot divorce the conjunctive
or combinatorial aspect from polyadic relations treated as bare intensions,
and analogously for properties as the limiting case. The relation of Exem-
plification (Instantiation or the like) is secondary and derivative, an artifact
of analysis, analogous to the predicates True and False, having a concep-
tual existence only (the medievals’ ‘second intentions’). In sum, the unity
between n entities does not always require an n+1 entity, the contrary be-
ing the road to regressive perdition. The latter error is abetted by the two
easy misidentifications of a) a predicate with its abstracted non-predicable
intension, and b) the vicious regress with the harmless progression of for-
mal exemplification instances. Rather, there can be a unification of exactly
n entities only if one of these entities has the categorical status of agent
unifier, and this is what we intuit to be the nature of predication, especially
as magnified in polyadic relations.

The above regress argument points to the proper understanding of

polyadic predication without the distorting mis-analogies of the ‘unsatu-
rated’ and ‘way’ metaphors, or a spatial interpretation of ‘between’. The
insight is: A polyadic predicate is an entity consisting of a qualitative con-
tent or intension controlling a combinatorial or linking state among a set of
relata. Properties as monadic predicates are the limiting case of polyadic
relations and thus have the same combinatorial character. Hence, all predi-
cates, monadic or polyadic, are ontoglial—each is a rigid linking holding
its relata both connected and yet separate, analogous to a connecting rod
between connected nodes. As noted, one sees this ‘relational’ conception
of predication in, for example, Aristotle’s construal of substantial forms
and their predicable structuring of plural matter in the Metaphysics. On the
combinatorial conception of predication three points are to be emphasized.
First, every predicate as predicable is a constituent of a complex—a fact—
that emerges as a function of the predicate’s unifying agency to or among
its subjects. In this predicates are the cause sine qua non of all plural unity,
i.e., of all complexity, organization, system, and structure, starting at the
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

lowest ontic level of atomic facts. Every pluralist ontology must appeal ul-
timately to relations, polyadic or monadic, in order to account for the unity,
however minimal, essential to every complex qua complex. This latter point

69
Bradley, Collected Essays, p. 638.

is crucial to ontology yet is generally neglected.70 Secondly, the intension

and the combinatorial state of a ‘relating relation’, i.e., in its predicable
role in a fact, are mutually necessitating. The intension is the source of
(i.e., is the cause of) the number, order, and compatibility/admissibility of
the linked relata, and, more globally, of the relation’s formal/logical prop-
erties. This is the more obvious the higher the adicity. For example, in the
fact :Between(a,b,c), which is the linear spatial arrangement of object b be-
ing between objects a and c, there is the unification of three objects under
the intension Between. The intension specifies the second-order attributes
of triadicity, the a—b—c order, and the logical properties such as the tri-
adic version of symmetry, i.e., if :Between(a,b,c) obtains then so does
:Between(c,b,a). Thirdly, the combinatorial agency of a predicate is ‘out-
wardly directed’, holding the predicate attached to but distinct from its sub-
ject(s). This is then equally true in the limiting monadic case—for fact
:P(a), the predicate with P as its intension attaches itself to a, it is not a
component of a. Significantly, then, all predicates are external to their sub-
jects in the sense that what is predicable of a subject (or subjects) is not a
constituent of the subject(s), contrary to the popular containment model of
predication. Note that this does not deny that a single subject a can have
predicates as constituents—the cross connecting relation instances making
up a as a complex are such predicates—but only that none of these predi-
cates can have emergent a itself as a relatum. Further, the externality of
predicates does not require that at some base level atomic predicates must
have bare particulars for subjects. Rather, the subjects of atomic predicates
need only be other atomic predicates, all of which are particulars on the ar-

70
Ontologies that attempt to avoid an appeal to relations as the agents of plural unity
resort to classes or mereological wholes taken as primitives. Yet, some explanation of
the source of unity of and correlative distinction among the elements of even these
minimal, bloodless wholes is required of an ontology. One cannot simply say that a
class or mereological sum exists when their elements exist, any more than one can say
a wall exists when the bricks exist. For everything that there is exists, including all
bricks across the world. Metaphorically, some entities and not others must be ‘brought
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

together’ to form the whole. Separating out the spatial mis-analogy, what this means is
that there must be a cause of the association of specific elements and in this differen-
tiation from all other groupings. The simple criterion of existence in its non-differ-
entiation could sustain at best the universal set. I propose that Russell was close to the
truth in observing that “Although from the point of view of a formal calculus one can
regard a relation as a set of ordered couples, it is the intension alone that gives unity to
the set. The same thing applies, of course, also to classes. What gives unity to the class
is solely the intension which is common and peculiar to its members.” Bertrand Rus-
sell, My Philosophical Development (London: Allen & Unwin, 1959), p. 67.

gument to follow. Moreover, the above does not imply that predicates are
radically external in the sense of their unifying agency being independent
of the natures or intensions of their subjects. Reiterating a point made pre-
viously, even the ‘external’ spatial relation of Between places conditions
an the possible categories of its relata, and is therefore not indifferent to
what are the contents or natures of the respective relata, e.g., it is predica-
tionally impossible for the triplet consisting of Plato, 3, and Circularity to
be subjects in any order for this relation. This incompatibility of intensions
is the source of what are termed ‘category mistakes’. Succinctly, then, un-
der the combinatorial analysis, the predication in fact :Rn(a1,a2,...,an) is a
rigid linking holding the non-identical relata among a1, a2, ..., an both
united and distinct, and where the respective natures of a1, a2, ..., an are
compatible with, and ordered (if any) according to, intension Rn.71

71
It is worth noting here that reinforcing this insight into the linking nature of relations
is its power in providing a viable ‘absence theory’ for solving the difficult problem of
true negative propositions. The intractability of this problem under standard ontology
is highlighted by George Molnar, ‘Truthmakers for Negative Truths’, Australasian Jour-
nal of Philosophy 78 (2000): 72-86. Atomic affirmative propositions, e.g., P(a), are held
to be true because they correspond to a ‘truth-maker’—fact :P(a). Some have held that,
by simple analogy, there must be for true negative propositions, e.g., ¬P(a), corre-
sponding ‘negative facts’. But this tack, in addition to positing dubious entities, does
not account for the fundamental syntactical fact that subjects cannot be negated, but
only linguistic predicates, e.g., whereas one can say ‘The ball is not red’, it is nonsen-
sical to say ‘The not ball is red’. On the theory of predication defended herein for a
negative proposition what is asserted is the absence among the referent(s) of the
grammatical subject term(s) (e.g., the ball) of a combinatorial/predicable state con-
trolled by the intension (e.g., Red) of the referent of the grammatical predicate. Here a
predicable nexus is absent, i.e., non-existent, whereas the controlling intension is ‘not
absent’ in the sense that it exists as an abstraction or construction independent of any
possible instantiation within a predicable instance. The fact that it is ontic predicates
that have this linking aspect explains why only the corresponding linguistic predicates
can be negated. Recently Hochberg has attempted to eliminate the need to posit nega-
tive facts by proposing that the truth-maker for a negative proposition ¬P(a) is the set
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

of all atomic facts corresponding to the true propositions of the form ‘P(a)’. In addi-
tion to other problems, this construction commits an epistemic fallacy of presumed
omniscience, a flaw that often infects the substitution of formal models for ontic enti-
ties. The set of atomic positive facts is to be the ground for the truth of the negative
proposition ¬P(a). Hence, to know that the negative proposition is true we would have
to survey the entire infinite set of facts in order to determine that the would-be fact
:P(a) is not in it. Since this is an impossible task for finite minds, we would never be
able to determine the truth of negative propositions. See Herbert Hochberg, ‘Facts and
Classes as Complexes and as Truth Makers’, The Monist 77 (1994): 170-91.

5. The Formal Distinction between a Predicate’s Combinatorial and

Intension Aspects

It is important to now observe the negative consequence of positing a real

pre-abstraction distinction between the combinatorial and intension aspects
of predicates and thus requiring a predicate to be a complex entity. Specifi-
cally, the unifier aspect of a predicate separated from any intension is a
‘bare linking’, i.e., a contentless or intensionless unifier, a notion as unten-
able as its cousin the ‘bare particular’.72 Yet, this is apparently what some
philosophers intend by the posit ‘non-relational ties’ as an answer to Brad-
ley’s Regress.

The trouble with bare unifiers is that, in having no intension to con-

trol or discriminate among possible subjects, such an entity would be noth-
ing but a blank association or arbitrary conjunction. A bare linking would
be the minimal and unconstrained unity of a list or set. Hence, if a bare
linking were the unifying agent within a fact, then there would be, for ex-
ample, no difference in the unification among R2, a, and b in the fact :R2(a,
b) and in the set {R2,a,b}. Thus having precisely the same components,
fact and set would have to be identical, which, for a contingent R2, would
require a set that exists necessarily to be identical with a contingent fact.
Moreover, a blank or ‘free’ association, in having no intension, specifies
no ordering among the entities linked, and hence despite, say, R2 being an
asymmetric or non-symmetric relation, there would be no difference be-
tween facts :R2(a,b) and :R2(b,a). For asymmetric or non-symmetric rela-
tion R2, it is R2-as-it-occurs-in-the-fact-:R2(a,b) i.e., as having the dual and
mutually influencing aspects of a delimiting intension and a combinatorial
state among a and b, that orders the latter. If the combinatorial nature is re-
assigned to a fourth constituent, the bare linking, then intension R2 be-
comes just one more relatum for an orderless triadic linking among R2, a,
and b, with no more influence over ordering the latter than it has in the set
{R2,a,b}. Finally and for the same reason, the modal character of every fact
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

becomes the same, i.e., all facts are necessary or all facts are contingent.
For, if the unity essential to every fact is the same bare unifier with no in-
tension Rn to control the modal attributes of these unions, then all such
complexes have the same modal quality, which is counterfactual. In short,
there are no ‘bare unifiers’ any more that there are ‘bare particulars’. To
anticipate, it is no accident that the arguments against bare unifiers and

72
See note 14.

bare particulars are analogous, for as will be seen presently the combinato-
rial aspect of a predicate as it exists among its subjects is unrepeatable, i.e.,
a particular.

Importantly then in regard to the Problem of Predication, what the

above demonstrates is that the combinatorial aspect of a predicate can exist
only with a correlative intension, and equally significant, it is inherent in
the nature of a predicate that there be no ‘ontic distance’ between these two
aspects, i.e., the predicate must be simple. For as seen, to make a predicate
internally complex by holding the unifier and intension aspects ontically
distinct components is to force the linking aspect itself to be devoid of con-
tent, and this is untenable. What this means is that the distinction between
these two aspects is ‘outside the thing’, i.e., they are abstractions represent-
ing partial realities not distinct in the predicate. This type of distinction is
what Duns Scotus identified as a distinctio formalis a parte rei (a formal
distinction on the side of the thing) and what other medievals referred to as
a distinctio rationis ratiocinatae (a distinction of the reasoned reason). Un-
der the designation ‘modal distinction’ the scholastic metaphysician Fran-
cisco Suarez draws this distinction in regard to the predicable ‘mode’ of a
property to its subject. In the face of Bradley’s regress, Suarez concludes
that “inherence does not need a further union or inherence by which it may
be united or may inhere, the reason is that the inherence does not add a real
entity which inheres and is united, but is merely a certain mode that of it-
self is the reason for union and inherence…. Yet this mode is not properly
distinct from the subject it modifies, as thing from thing.”73 This is a dis-
tinction of the reasoned reason where “Although the same object is appre-
hended in each concept, the whole reality contained in the object is not
adequately represented, nor is its entire essence and objective notion ex-
hausted, by either of them…. Things said to be thus distinct are real enti-
ties, or rather a single real entity conceived according to various aspects.”74
Recently, Armstrong has concurred with Scotus, saying, “It is concluded,
therefore, that although particularity and universality are inseparable as-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

pects of all existence, they are neither reducible to each other nor are they
related. Though distinct, their union is closer than relation.”75 If the unifier
and particularizer of the atomic ontic unit are the same, as will be argued

73
Suarez, On the Various Kinds of Distinctions, I. 18-19, pp. 29-31.
74
Ibid., I. 5, p. 19.
75
D. M. Armstrong, A Theory of Universals: Universals & Scientific Realism, Vol. II
(Cambridge: Cambridge University Press, 1978), p. 3.

presently, then Campbell’s advocacy of the formal distinction is here apro-

pos: “Although the idea here is that all particulars are particulars, and each
of them has a nature, this does not involve conceding that a trope is after
all complex (a union of particularity with a nature-providing property). The
distinction is perhaps a ‘formal’ one, as Scotus used the term.”76 Against
the formal distinction, Moreland has recently reiterated a position main-
tained by Ockham that if what is distinguished in thought are aspects of a
single entity then they must be distinct within the entity.77 This is an asser-
tion, not an argument, and, to the contrary, there is no logical necessity
preventing a single subject entity from being rich enough to support sepa-
rate and partial yet veridical and complementary conceptions abstracted
from it. This type of distinction is logically possible and necessitated on
the above analysis.78

6. The Individuation of Predicates

Now, with this ‘non-complex’ character of predicates as a premise, it is

possible to demonstrate that predicates are individuated as instances, Rni.
Specifically, it is a simple matter to show that the combinatorial aspect of a
predicate is unrepeatable, and this together with the subsuming predicate’s
internally simple nature implies the latter’s inherited unrepeatability, i.e.,
individuation. In this we have the principium individuationis that eludes all
bundle, substratum, and individuation-by-instantiation theories, and pro-
vides an explanation for what nominalists take as a primitive datum. The
argument can be synopsized as follows.79 Let it be the case that two facts
:Rn(a,b) and :Rn(b,a) obtain, where Rn is a contingent and, for present pur-
poses, a non-symmetric relation intension, e.g., Love. As non-symmetric,
Rn imposes an order on its mutual combinatorial aspect accounting for the
directional unity in each fact. But this directional unity in not identical
across the two facts—in :Rn(a,b) the predicational linking is from-a-to-b

77
Moreland, Universals, Qualities, and Quality-Instances, p. 68. Also his ‘Keith Camp-
bell and the Trope View of Predication’, Australasian Journal of Philosophy 67 (1989):
386-7; William of Ockham, Ordinatio, d. 2, q. 6, trans. by Paul Spade, Five Texts on the
Mediaeval Problem of Universals (Indianapolis: Hackett Publishing Co., 1994), pp. 153-
90.
78
See Mertz, ‘Individuation and Instance Ontology’, for an analogy to show the possi-
bility of a simple entity with dual abstractable aspects.
79
An expanded argument is found in my ‘Individuation and Instance Ontology’ [and
in other essays herein].

whereas in :Rn(b,a) the predication is from-b-to-a. If the predicable union

were numerically identical in both facts, then having exactly the same con-
stituents—intension Rn, relata a and b, and identical union—then by the
Uniqueness by Composition Principle the facts would be identical, :Rn(a,b)
= :Rn(b,a). This would follow for all relata pairs of Rn, and hence Rn would
be symmetric rather that non-symmetric. Further, no matter what the formal
properties of intension Rn are (e.g., whether symmetric, asymmetric, or non-
symmetric), on the assumption that Rn is contingent, if it were numerically
the same fact-forming-union in fact :Rn(a,b) and any other fact :Rn(c,d), then
if one fact ceased to obtain, i.e., its predicational union ceased to exist, then
similarly all other like-intensioned facts would cease to exist. That is, all
such facts with contingent Rn would come into and go out of existence to-
gether. All of this is a reductio of the present assumption that it is numeri-
cally the same predicable nexus in facts having the same intension. It fol-
lows that the combinatorial union of a predicate with its subject n-tuple is
unique and unrepeatable relative to that n-tuple. Because this state of unifi-
cation is individuated, and because it and its correlative intension Rn are
distinct aspects in abstraction from but not distinct within the subsuming
simple predicate, this predicate is itself an individuated instance Rni. In
short, the predicable nexus is not repeatable, and hence a predicate qua
predicable is not a universal.

It is the case, then, that predication is the principle or cause of indi-

viduation. Hence, contrary to a sometimes proposed thesis that it is the in-
dividuation of subjects that externally individuates their attributes, on the
above an n-tuple of subject relata, <a1,a2,...,an>, is the secondary cause of
the individuation of a predicate Rni, but it is the internal combinatorial as-
pect of Rni among these relata, what is essential to and inseparable from
fact :Rni(a1,a2,...,an), that is the primary cause of the predicate’s individua-
tion. Unlike for trope theory, where for exactly resembling tropes it is a
brute fact that they differ solo numero, we have here the explanation that
relation instances differ by their individuated linking among distinct relata
sets. Further, because the individuator of a predicate Rni is its combinatorial
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

state, which is not identical with its correlative intension Rn and not itself a
further intension, we do not have the absurd situation where the same en-
tity, instance Rni, has contradictory attributes, i.e., unrepeatability and re-
peatability. Nor do we have as with tropes the mysterious compression of
individuation and intension into a simple entity that somehow removes the
repeatability of the latter.

7. Conclusion: Network Instance Realism

The above results provide a successful, non-reductive means for solving

ontology’s fundamental Triple Aspect Problem and the Problem of Com-
plexity, and those problems that, through the pivotal role of predication,
are attendant to them.80 With the single category of relation instances we
have simple entities that are of their essence predicable, i.e., combinators,
that by that very fact are individuated, and where each has a correlative
non-predicable repeatable intension. Predicable particulars, some neces-
sary, some contingent, along with their abstractable intensions, are suffi-
cient to account for all of observable reality and conform to our best scien-
tific explanations thereof. The resulting ontology would outline as follows.

First, as outwardly combinatorial, atomic instances are dependent for

their existence upon other entities, but this does not make them less real
(‘supervenient’) or incapable of sustaining a complete ontology. The status
of ontic dependence of a relation instance on its relata, whether further
characterized as necessary (an ‘internal’ relation) or as contingent (an ‘ex-
ternal’ relation), does not militate against the distinct reality of the instance
anymore than does, say, the existential dependence of a cat upon its food
make the cat ‘less real than’ or ‘supervene upon’ its sustenance. Indeed, the
forgoing displays how from among the traditional characterizations of sub-
stance the condition of being independent (‘self-sustaining’) conflicts with
that of being a cause of unity. Guided by the criterion for being real sug-
gested by the Stranger in Plato’s Sophist (247e), i.e., what is real has the
power (i.e., active and passive dynamis) to affect and be affected by some-
thing else, this paraphrases on the above into the more general form: An
entity is real that has the active power to go beyond itself and ‘make a dif-
ference’ in, i.e., qualify, something else, or the passive power to be so
qualified. The former power is that of predicates, especially in the broad
80
The problem solving power of this analysis of individuated predicates with universal
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

intensions is further magnified when the logic inherent in it is formalized. The result is
a ‘formal ontology’ in the sense of Nino Cocchiarella, ‘Ontology II: Formal Ontology’
in Handbook of Metaphysics and Ontology, eds. H. Burkhardt & B. Smith (Munich:
Philosophia Verlag, 1991), pp. 640-47. The logic inherent in realist instance ontology,
what I have called ‘PPL’, has a number of powerful results, e.g., distinguishing legiti-
mate from illegitimate self-referential predication, distinguishing identity from indis-
cernibility, and providing an ontology for arithmetic. PPL is developed in Moderate
Realism, with an improved version in ‘The Logic of Instance Ontology’, Journal of Phi-
losophical Logic 28 (1999): 81-111[Essay 6 herein].

polyadic sense and as conditioned by intensions, and the latter power is

that of subjects, again as conditioned by their contents. This and the unre-
ducible/uneliminable status of relations places them among the fundamen-
tally real, and for all controlling intensions, not just those artificially re-
stricted to spatial/temporal, efficient causal, or (contra Armstrong) contin-
gent types. All of this with the further fact of their individuation renders
predicates of any n-adicity the primary ontic category. Starting at the
atomic ontic level, relation (including property) instances would have re-
lata from among themselves, forming lattice-like complexes. [For details
and models see Essays 2 and 4 in this volume.] One can model this possi-
bility physically. In this we have the answer to philosophers (e.g., Campbell
and Armstrong) who contend that if there were no non-relations (i.e., non-
predicable entities—‘objects’, ‘substances’, etc.) there would be no relations,
but not vice versa. We see that exactly the opposite is the case. What atomic
relations there are (concrete or abstract (including logical/mathematical)) is
for science in the broad sense to determine. These atomic lattices as unit
wholes function themselves as single relata for further relation instances, re-
sulting in complexes of complexes, and so on up through hierarchies of in-
creasingly intricate wholes, static and dynamic (i.e., as ‘event structures’).

It is to be emphasized that a complex is not definable simply as any

set or ‘conjunction’ of relation instances, but rather as a set of instances
that are themselves ‘connected’ by means of strategically shared relata.
That is, in its full generality a complex is a network of relation instances
where, as such, each instance is joined to one or more other instances via
shared relata nodes, all in such a way as to yield continuous connectivity,
analogous to that facilitated by the sharing of vertices in a lattice. Simons has
made this point precise using the topological analog of path-connectedness.81
Loosely, two relata a and b are path-connected if and only if through a
chain of relations and shared relata one can trace a path linking a and b. A
complex is then a set of relations and relata where any two of the latter are
path-connected through chains of the former. Alternately and with a com-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

bined precision and simplicity made possible with the refinement of in-
stances, we can define a set of instances to be a complex if and only if any
two disjoint and exhaustive subsets of it have at least one instance each
that share the same relatum. The accuracy of this definition is verified by
means of simple connected ‘road and node’ diagrams.

81
Peter Simons, Parts: A Study in Ontology (Oxford: Clarendon Press, 1987), p. 327.

Complexes at any level are non-predicable entities (‘objects’ in the

sense of Frege) and relatively independent (in the manner of traditional
substances), but both their constituent instances and the complexes as
wholes can be possible relata for further relation instances. In this ontology
we have the unity-among-the-diverse requisite of a plural and multiply
structured universe given in experience and theorized in science. Conso-
nant with modern science (e.g., neural science and quantum mechanics)82,
at different levels n-adic relations emerge each with its own novel inten-
sion, and each non-existent at, and hence not reducible to, lower levels of
complexity. At certain levels of neural complexity cognitive faculties arise
sufficient for the then emergence of instances of logical, mathematical,
ethical, emotional, etc., relations. At these levels of complexity there also
arise both the facility to abstract the content intensions from relation in-
stances and the facility to freely construct from them further intensions that
may or may not be instantiated (this the basis for fantasy, theory, and the
modal character of possibility). This possibility of intensions without
predicable instances runs contrary to a popular principle that holds that
every n-adic universal must have at least one n-tuple of entities that instan-
tiate it. Uninstantiated intensions are the basis for an actualism that allows
‘possible worlds’ as conceptual constructs only—the possible is delimited
by intensions, the actual by intensions that have instances. Over all, some
intensions emerge only at the conceptual level and of these some are in-
stantiated (e.g., True/False, the logical connectives, Exemplification,
mereology’s Part-of) whereas others are not (e.g., Unicorn, Phlogiston),
while other intensions exist extra-conceptually as the content of instanti-
ated instances (e.g., instances of space-time relations, or of causal rela-
tions). It is not, as some philosophers claim, just spatial and/or causal rela-
tions that form the ‘cement of the universe’, but rather instances of rela-
tions of every kind of intension. In general, because of the inter-relatedness
under various intensions of all entities, whether transcendent or natural
(i.e., a relatum in the space-time matrix), reality is to be characterized as
the Total-Complex, this in contrast to the distinction-obliterating homoge-
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

neous One of monism.

Contra historically popular substance ontology, no natural entities

exist per se; rather they exist as, or as abstracted or constructed intension
82
On the need for emergent properties in science see Michael Silberstein and John Mc-
Geever, ‘The Search for Ontological Emergence’, The Philosophical Quarterly 49 (1999):
182-200.

nodes in, substructures in the all-encompassing physical/cognitive complex

that is the natural world and that is sustained at every level inter se. The
traditional ‘substance’ is here a sub-complex that as a unit endures through
the ‘accidental’ changes of becoming or ceasing to be a relatum for relation
(including property) instances not constitutive of (‘outside’) the complex,
but that suffers ‘substantial change’ when one or more of the instances in-
ternal (‘essential’) to the complex cease to obtain and are not replaced by
instances of the same kind, thus causing this complex as a whole to cease
to exist. This is in contrast to such devices as Simons’ theory of substratum
‘kernels’ of ‘mutually founding’ tropes that form the essence of the sub-
stance and around which there is a lesser bundle of accidental tropes.83 In
network instance realism ‘substance’ complexes are controlled and delim-
ited by the internal real intensions of the constituent relation instances
where a few of all possible n-tuples are related, this being opposed to the
mis-identification of ‘substances’ with certain mereological sums84 chosen
from among a formal pre-existing set of sums, one for every n-tuple, the
latter posited in a misguided attempt at avoiding ontological commitment.85

83
Simons, ‘Particulars in Particular Clothing’.
84
David Lewis has in his Parts of Classes advocated the view that mereology provides
a general theory of composition. Against this view Simons has pointed out the limita-
tions of mereology in explicating integrated wholes. See Simons, Parts: A Study in
Ontology, pp. 324ff.
85
Mereological sums are to be the composites of any disjoint objects whatsoever, though
ontically no more than the member objects. Yet the theory prescinds from any account of
the unification of the elements, a defect of omission seen even in the etymology of
‘compose’, i.e., ‘to place together’, and in this masks what is the ontic necessity of
both an ‘agent of unification’ and a resultant whole distinct from all the elements. In
addition, the actual universe is a selectively ordered and in part dynamically structured
whole narrowly delimited as such from among all possible structurings by the required
mutual compatibility of natures (i.e., the intensions) of the linking instances and their
relata. Mereology, by contrast, posits for every fixed universe of disjoint entities what-
soever what would be a complete and static monolithic formal structure consisting of
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

all possible sums and subsequent substructures under the Part-of relation. Modeling in
mereology is selecting out from, i.e., ‘subjectively creating’, one formal substructure
from what is in itself the total and undivided universe structured under the Part-of rela-
tion. There is nothing in this total mereological universe itself to distinguish one sub-
structure from another. A substructure is an artifact of human practice (a ‘fiat object’)
selected because of its fit with a prior structure of the experienced and theorized real.
Mereology is a useful formal fiction but it is not ontology, and it is for this reason and
not for what some philosophers (e.g., Armstrong) hold—because they ‘supervene’
upon their parts—that mereological sums have no ontological standing over and above

The equally important category of events, rather than being misidentified

with artificial constructions, e.g., set-theoretical n-tuples of n subjects, a
relation, and a time, are here explicable as relation instances. Processes are
then seen as instance complexes having substructures consisting of sequen-
tially related temporal or causal relation instances.86

In sum, on the above analysis the ultimate categorical structure of re-

ality is the simple dichotomy of predicable and non-predicable entities.87
Predicable entities are primary and are relation (including property) in-
stances. Non-predicable entities are secondary and are either complexes of
relation instances or intension universals abstracted from these instances.
This accounts for the world as a ‘totality of facts, not things’, and in its ob-
served and theorized form of rich interdependence and layered emergence.
A world of hierarchical complexes requires a relational/structural ontology,
and it is the correlative combinatorial and individuating aspects of predi-
cates, writ-large in polyadic predicates, that make this ontology possible
and intuitive.
Copyright © 2015. De Gruyter, Inc.. All rights reserved.

their parts. The same critique applies to set theory, mutatis mutandis. See note 70
above and Moderate Realism, pp. 51-58.
86
For a description of the advantages of instance ontology in the analysis of events and
the logic of causation see Moderate Realism, pp. 78-80.
87
Characterizing the ontology advocated here as ‘two-category’ corrects its less accurate
characterization in Moderate Realism as a ‘one-category’ ontology. Though the one cate-
gory of predicable relation instances is primary, derivative entities in the category of non-
predicable entities, complexes and universals, are equally real, though sufficiently distinct
as to form two sub-categories.

An Introduction To Metametaphysics
No ratings yet
An Introduction To Metametaphysics
270 pages
A Most Unlikely God
100% (1)
A Most Unlikely God
93 pages
Alain Badiou Being and Event PDF
83% (6)
Alain Badiou Being and Event PDF
272 pages
Badiou Being and Event 1988
91% (11)
Badiou Being and Event 1988
525 pages
Higgins - Notes On Categories and Groupoids
No ratings yet
Higgins - Notes On Categories and Groupoids
195 pages
Wesley Phoa - Fibrations, Toposes, Set Theory
No ratings yet
Wesley Phoa - Fibrations, Toposes, Set Theory
152 pages
Against "Ostrich" Nominalism 1980
No ratings yet
Against "Ostrich" Nominalism 1980
10 pages
Essays On Realist Instance Ontology And Its Logic Donald W Mertz download
No ratings yet
Essays On Realist Instance Ontology And Its Logic Donald W Mertz download
75 pages
Cocch Rome5
No ratings yet
Cocch Rome5
22 pages
3 The Nature and Necessity of Composite Simples, E.g., Ontic Predicates
No ratings yet
3 The Nature and Necessity of Composite Simples, E.g., Ontic Predicates
42 pages
25462505
No ratings yet
25462505
78 pages
On Ontology and Realism in Mathematics
No ratings yet
On Ontology and Realism in Mathematics
36 pages
K. Mulligan Language, Truth and Ontology Philosophical Studies Series 1991
100% (1)
K. Mulligan Language, Truth and Ontology Philosophical Studies Series 1991
223 pages
Management 225 Chapter 6 Worksheet
No ratings yet
Management 225 Chapter 6 Worksheet
6 pages
Logic Essenceand Modality
No ratings yet
Logic Essenceand Modality
28 pages
Mind Association, Oxford University Press Mind
No ratings yet
Mind Association, Oxford University Press Mind
23 pages
The Paradox of Identity
No ratings yet
The Paradox of Identity
14 pages
Predication and The Problem of Universals
No ratings yet
Predication and The Problem of Universals
27 pages
Armstrong - Universals
100% (1)
Armstrong - Universals
82 pages
Ontologia SEP
No ratings yet
Ontologia SEP
308 pages
Elaine Huang metaphysics Paper1
No ratings yet
Elaine Huang metaphysics Paper1
7 pages
Landini 1991
No ratings yet
Landini 1991
20 pages
Paper 1
No ratings yet
Paper 1
4 pages
51971
No ratings yet
51971
77 pages
Ontology and Ideology
No ratings yet
Ontology and Ideology
5 pages
Formal Ontology And Conceptual Realism Cocchiarella Nino instant download
100% (3)
Formal Ontology And Conceptual Realism Cocchiarella Nino instant download
89 pages
Universals, Communicable Knowledge, and Metaphysics - Charles A. Baylis
100% (1)
Universals, Communicable Knowledge, and Metaphysics - Charles A. Baylis
10 pages
The Correspondence Theory of Truth PDF
No ratings yet
The Correspondence Theory of Truth PDF
14 pages
Semantic Singularities: Paradoxes of Reference, Predication, and Truth Keith Simmons - The full ebook with complete content is ready for download
100% (1)
Semantic Singularities: Paradoxes of Reference, Predication, and Truth Keith Simmons - The full ebook with complete content is ready for download
51 pages
Alain Badiou Being and Event PDF 1 136
No ratings yet
Alain Badiou Being and Event PDF 1 136
136 pages
Metaphysics and Scientific Realism PDF
100% (1)
Metaphysics and Scientific Realism PDF
270 pages
The_Genetic_Origination_of_Truth_Toward_Being,_Edith_Stein’s_Reconfiguration
No ratings yet
The_Genetic_Origination_of_Truth_Toward_Being,_Edith_Stein’s_Reconfiguration
113 pages
Week 1: W/Aug 26 Introduction
No ratings yet
Week 1: W/Aug 26 Introduction
3 pages
Australia Philosophy Convergence 2024
No ratings yet
Australia Philosophy Convergence 2024
30 pages
Quantification Transcending Beyond Frege
No ratings yet
Quantification Transcending Beyond Frege
249 pages
(Ebook) Universality and the Liar: An Essay on Truth and the Diagonal Argument by Keith Simmons ISBN 9780511551499, 9780521061391, 9780521430692, 0511551495, 0521061393, 0521430690 download
100% (1)
(Ebook) Universality and the Liar: An Essay on Truth and the Diagonal Argument by Keith Simmons ISBN 9780511551499, 9780521061391, 9780521430692, 0511551495, 0521061393, 0521430690 download
60 pages
Zalamea - Peirce's Continuum - Part 1
100% (1)
Zalamea - Peirce's Continuum - Part 1
46 pages
Hochberg 2013 - Nominalism and Idealism
No ratings yet
Hochberg 2013 - Nominalism and Idealism
22 pages
Levinson PhilosophicalReview 1992
No ratings yet
Levinson PhilosophicalReview 1992
8 pages
Hacking, Ian - On the Reality of Existence and Identity (1978)
No ratings yet
Hacking, Ian - On the Reality of Existence and Identity (1978)
20 pages
Universals - Chad Carmichael
No ratings yet
Universals - Chad Carmichael
18 pages
Instant Download Universality and the Liar An Essay on Truth and the Diagonal Argument 1st Edition Keith Simmons PDF All Chapters
100% (4)
Instant Download Universality and the Liar An Essay on Truth and the Diagonal Argument 1st Edition Keith Simmons PDF All Chapters
71 pages
Semantic Singularities: Paradoxes of Reference, Predication, and Truth Keith Simmons download
100% (2)
Semantic Singularities: Paradoxes of Reference, Predication, and Truth Keith Simmons download
53 pages
Projective Geometry As A Model For Hegel
No ratings yet
Projective Geometry As A Model For Hegel
20 pages
Revitalizing What He Takes To Be The Perennial Core of Such Questions. So Although It Might Seem Like A
No ratings yet
Revitalizing What He Takes To Be The Perennial Core of Such Questions. So Although It Might Seem Like A
25 pages
Inwagen Ontological Arguments
No ratings yet
Inwagen Ontological Arguments
23 pages
Stipulations_Missing_Axioms_in_Frege_s_Grundgesetze_der_Arithmetik
No ratings yet
Stipulations_Missing_Axioms_in_Frege_s_Grundgesetze_der_Arithmetik
37 pages
Universals An Opinionated Introduction by D. M. Armstrong
No ratings yet
Universals An Opinionated Introduction by D. M. Armstrong
158 pages
The Immanence of Truths _ Being and Event III -- Alain Badiou; Susan Spitzer; Kenneth Reinhard -- Bloomsbury UK, London [England], 2021
No ratings yet
The Immanence of Truths _ Being and Event III -- Alain Badiou; Susan Spitzer; Kenneth Reinhard -- Bloomsbury UK, London [England], 2021
625 pages
Menzel - Pure Logic and Higher-Order Metaphysics
No ratings yet
Menzel - Pure Logic and Higher-Order Metaphysics
42 pages
anellis2012
No ratings yet
anellis2012
13 pages
Object A4
No ratings yet
Object A4
27 pages
Mactao-3 1
No ratings yet
Mactao-3 1
29 pages
The Place of Logic in Philosophy
No ratings yet
The Place of Logic in Philosophy
4 pages
Duke University Press Philosophical Review
No ratings yet
Duke University Press Philosophical Review
26 pages
THEORIA. Revista de Teoría, Historia y Fundamentos de La Ciencia 0495-4548
No ratings yet
THEORIA. Revista de Teoría, Historia y Fundamentos de La Ciencia 0495-4548
19 pages
The Relativity of Everything: A Philosophical Inquiry into Relativity, Complexity, and Distributed Society
From Everand
The Relativity of Everything: A Philosophical Inquiry into Relativity, Complexity, and Distributed Society
Rodney Morris
No ratings yet
Treatise on the Neurophilosophy of Consciousness: A Multidisciplinary Biopsychosocial (Bps) Model
From Everand
Treatise on the Neurophilosophy of Consciousness: A Multidisciplinary Biopsychosocial (Bps) Model
Dr. Angell O. de la Sierra
No ratings yet
The Limits of Knowledge
From Everand
The Limits of Knowledge
Paul O'Hara
No ratings yet
Hawking Radiation 4
From Everand
Hawking Radiation 4
Roman Andie
No ratings yet
Projection: New Terminology Based on a Dual System Derived from the Chinese Book of Changes and Applied to Psychological Considerations of Self, Action and Influence
From Everand
Projection: New Terminology Based on a Dual System Derived from the Chinese Book of Changes and Applied to Psychological Considerations of Self, Action and Influence
James P. Devaney
No ratings yet
Relational Gestalt Practice: Key Concepts
From Everand
Relational Gestalt Practice: Key Concepts
The Gestalt Legacy Project
4/5 (2)
Comments on “A Bio-Cultural-Historical Approach to the Study of Development (2016)”
From Everand
Comments on “A Bio-Cultural-Historical Approach to the Study of Development (2016)”
Razie Mah
No ratings yet
Microgenetic Theory and Process Thought
From Everand
Microgenetic Theory and Process Thought
Jason W. Brown
No ratings yet
PSMLM - Demonic Temptations
No ratings yet
PSMLM - Demonic Temptations
31 pages
Esakia - Intuitionistic Modal Logic Via Topology
No ratings yet
Esakia - Intuitionistic Modal Logic Via Topology
16 pages
Plotki and Turi - Towards_a_mathematical_operational_semantics
No ratings yet
Plotki and Turi - Towards_a_mathematical_operational_semantics
12 pages
Brown - Groupoid Survey
No ratings yet
Brown - Groupoid Survey
27 pages
Longo and Moggi - CCcateg-for-types
No ratings yet
Longo and Moggi - CCcateg-for-types
22 pages
Nicolas Vaughan - Ockham on Logic as a Rational Science (D.Phil Thesis)
No ratings yet
Nicolas Vaughan - Ockham on Logic as a Rational Science (D.Phil Thesis)
294 pages
Wolenski - Semantics-And-Truth
No ratings yet
Wolenski - Semantics-And-Truth
392 pages
Knuttila - Roger Roseth and Medieval Deontic Logic
No ratings yet
Knuttila - Roger Roseth and Medieval Deontic Logic
14 pages
Tech-Salary-Guide 2021 NorthAmerica PDF
No ratings yet
Tech-Salary-Guide 2021 NorthAmerica PDF
18 pages
LF I and Dacosta Systems
No ratings yet
LF I and Dacosta Systems
28 pages
Scotus On Identity... I Mean, Mereological Fusions
No ratings yet
Scotus On Identity... I Mean, Mereological Fusions
35 pages
Ockham On Priority and Posteriority: Abstract
No ratings yet
Ockham On Priority and Posteriority: Abstract
34 pages
Richard of Conington
No ratings yet
Richard of Conington
16 pages
A Theory Of Minimalism Marc Botha download
No ratings yet
A Theory Of Minimalism Marc Botha download
74 pages
Fe 111 Philosophy of Education. i . Module-1
No ratings yet
Fe 111 Philosophy of Education. i . Module-1
81 pages
Curriculum Traditions
No ratings yet
Curriculum Traditions
45 pages
Epistemological Background of Natural Law
No ratings yet
Epistemological Background of Natural Law
18 pages
Modern Legal REALIST Outline
No ratings yet
Modern Legal REALIST Outline
4 pages
Complete Download The Idea of International Society Erasmus Vitoria Gentili and Grotius Ursula Vollerthun PDF All Chapters
100% (1)
Complete Download The Idea of International Society Erasmus Vitoria Gentili and Grotius Ursula Vollerthun PDF All Chapters
65 pages
Secondary School Curriculum Instruction
100% (1)
Secondary School Curriculum Instruction
137 pages
ST1 Notes
No ratings yet
ST1 Notes
18 pages
Panion To The Philosophy of Science 2011 RETAIL eBook-ZOiDB00K PDF
100% (3)
Panion To The Philosophy of Science 2011 RETAIL eBook-ZOiDB00K PDF
465 pages
Download A Plea for Natural Philosophy: And Other Essays Penelope Maddy ebook All Chapters PDF
100% (1)
Download A Plea for Natural Philosophy: And Other Essays Penelope Maddy ebook All Chapters PDF
47 pages
Jurisprudence Dec 2018 Sem 3
No ratings yet
Jurisprudence Dec 2018 Sem 3
83 pages
Six Principles of Political Realism
86% (7)
Six Principles of Political Realism
2 pages
Download Full (Ebook) Qualitative Research Practice: A Guide for Social Science Students and Researchers by Jane Beaglehole Ritchie; Jane Lewis; Carol McNaughton Nicholls; Rachel Ormston ISBN 9781446209127, 1446209121 PDF All Chapters
100% (4)
Download Full (Ebook) Qualitative Research Practice: A Guide for Social Science Students and Researchers by Jane Beaglehole Ritchie; Jane Lewis; Carol McNaughton Nicholls; Rachel Ormston ISBN 9781446209127, 1446209121 PDF All Chapters
86 pages
Infosys ps-2
No ratings yet
Infosys ps-2
15 pages
Sameness and Substance Renewed - David Wiggins
100% (1)
Sameness and Substance Renewed - David Wiggins
274 pages
History, Meaning, and Interpretation: A Critical Response To Bevir
No ratings yet
History, Meaning, and Interpretation: A Critical Response To Bevir
12 pages
Traits Alport
No ratings yet
Traits Alport
182 pages
Essay I by Nicolas Laos PDF
No ratings yet
Essay I by Nicolas Laos PDF
13 pages
Moore (Themes - From - G. - E. - Moore)
No ratings yet
Moore (Themes - From - G. - E. - Moore)
359 pages
BPEd 101 102
No ratings yet
BPEd 101 102
52 pages
Dennis Patterson - Normativity and Objectivity in Law PDF
No ratings yet
Dennis Patterson - Normativity and Objectivity in Law PDF
40 pages
Arguments Concerning Naturalism
No ratings yet
Arguments Concerning Naturalism
35 pages
Abstract - The Teaching of Basketball As A Process of Educational Subjectification - Bel
No ratings yet
Abstract - The Teaching of Basketball As A Process of Educational Subjectification - Bel
4 pages
A Realist Conception Of Truth William P Alston pdf download
No ratings yet
A Realist Conception Of Truth William P Alston pdf download
83 pages
Full download (Ebook) Worlds and Individuals, Possible and Otherwise by Takashi Yagisawa ISBN 9780199576890, 0199576890 pdf docx
No ratings yet
Full download (Ebook) Worlds and Individuals, Possible and Otherwise by Takashi Yagisawa ISBN 9780199576890, 0199576890 pdf docx
81 pages
Philos of Education
No ratings yet
Philos of Education
47 pages
Technology Ethics A Philosophical Introduction And Readings Gregory J Robson pdf download
100% (2)
Technology Ethics A Philosophical Introduction And Readings Gregory J Robson pdf download
81 pages
Instant Download Studies in Philosophical Realism in Art Design and Education 1st Edition Neil C. M. Brown (Auth.) PDF All Chapters
100% (1)
Instant Download Studies in Philosophical Realism in Art Design and Education 1st Edition Neil C. M. Brown (Auth.) PDF All Chapters
55 pages