1
LEIBNIZ ON COMPOSSIBILITY
James Messina (UCSD) and Donald Rutherford (UCSD)
Abstract
Leibniz’s well-known thesis that the actual world is just one among many
possible worlds relies on the claim that some possibles are incompossible,
meaning that they cannot belong to the same world. Notwithstanding its central
role in Leibniz’s philosophy, commentators have disagreed about how to
understand the compossibility relation. We examine several influential
interpretations and demonstrate their shortcomings. We then sketch a new
reading, the cosmological interpretation, and argue that it accommodates two key
conditions that any successful interpretation must satisfy.
At the heart of Leibniz’s philosophy is the claim that the actual world exists contingently, as just
one among many possible worlds. God has chosen to create this world because it is the best, but
God could have chosen differently. Leibniz opposes his position to that of Spinoza, according to
whom everything that could exist does exist and nothing could have existed differently than it
does.1 Leibniz regards Spinoza’s denial of contingency as a conclusion to be avoided at all
costs.2 In order to rebut it, he must explain at a minimum why not all possibles are actual.3 To
this end, Leibniz invokes the notion of compossibility:
[N]ot all possibles are compossible. Thus, the universe is only a certain collection
of compossibles, and the actual universe is the collection of all existing possibles,
that is to say, those which form the richest composite. And since there are
2
different combinations of possibilities, some of them better than others, there are
many possible universes, each collection of compossibles making up one of them.
(GP III 573/L 662)
Not all possibles are actual, because not all possibles are compossible with the set of actual
substances. Those possibles that are compossible belong to the same possible world. Thus, the
compossibility relation explains why some substances are merely possible and partitions possible
substances into different worlds, among which God chooses the best.4
While there is general agreement about the role Leibniz assigns to the compossibility
relation, there is no consensus about its basis. Commentators have distinguished two main ways
of explicating Leibniz’s view. On the logical interpretation, two substances are compossible if
and only if the supposition of their joint existence is logically consistent.5 On the lawful
interpretation, two substances are compossible if and only if they are suitably related under some
set of universal laws.6 One of the chief issues dividing these accounts is whether a failure of
compossibility, i.e. incompossibility, implies a logical restriction on the exercise of God’s power,
or whether it is a result merely of certain free decrees that God considers in organizing
substances into worlds. On the logical interpretation, it is impossible for God to actualize two
incompossible substances, for the existence of one logically excludes the existence of the other.
The lawful interpretation posits no such restriction on the exercise of God’s power; rather, it
assumes simply that in conceiving of a world God chooses to consider only those substances
among which certain lawful relations would be observed.
In what follows, we provide an overview of the debate concerning Leibniz’s doctrine of
compossibility and propose a novel resolution of it. In sections I and II, we examine the logical
and lawful interpretations, explain their respective motivations and document their shortcomings.
3
In section III, we consider two hybrid readings combining features of the logical and lawful
interpretations, and show that these too are unsatisfactory. From this analysis, we extract two
requirements, the Connection and Inclusion Conditions, which we argue any satisfactory
interpretation of compossibility must satisfy. In section IV, we sketch a new account, the
cosmological interpretation, which we claim meets these conditions while avoiding the problems
faced by its rivals.
I
In his influential discussion of compossibility and possible worlds, Benson Mates draws on
Leibniz’s complete concept theory to explain the nature of a possible world. According to
Leibniz, each possible substance is represented by a concept that is complete, insofar as it entails
everything that would be truly predicated of that substance were it to exist (A VI.iv 1540/AG 41;
cf. A VI.iv 1546-49/AG 44-6).7 Given Leibniz’s principle of the identity of indiscernibles—that
there cannot be numerically different, but qualitatively indistinguishable things—it follows that
each substance is associated with a unique complete concept that is sufficient to distinguish it
from every other possible substance. Mates defines a possible world as any maximal consistent
set of complete concepts (‘Leibniz on Possible Worlds’ 340; The Philosophy of Leibniz 75).8 He
thus regards compossibility as a relation that holds between complete concepts. The fact that
two possible substances are or are not members of the same world is fully determined by the
content of their complete concepts.
For Mates, like others who espouse a logical construal of compossibility, the supposition
of the joint existence of two incompossible substances involves a logical contradiction: ‘A pair
of individual concepts, A and B, are compossible if no contradiction follows from the supposition
4
that there are corresponding individuals for both of them—that is, if the statements “A exists”
and “B exists” are consistent with one another’ (The Philosophy of Leibniz 75). Such a reading is
not without textual support, but it is less extensive than sometimes supposed.9 The principal
challenge for this account is to explain how complete concepts can exclude one another
logically. The problem is especially acute if one maintains, as Mates does, that the complete
concept of a substance includes only its ‘simple’ properties, i.e., those that are both non-
relational and not analyzable in terms of more basic properties (‘Leibniz on Possible Worlds’
339-40; The Philosophy of Leibniz 219-20). Leibniz himself highlights the problem in a well-
known text: ‘It is as yet unknown to men whence arises the incompossibility of different things,
or of how different essences can conflict with each other, since all purely positive terms seem to
be compatible with each other’ (GP VII 194).
Nicholas Rescher’s solution to this puzzle is to abandon the view that a complete concept
includes only ‘simple’ or ‘purely positive’ properties. If complete concepts include relational
properties in addition to monadic ones, then it is possible to make sense of a contradiction arising
from the supposition of the joint instantiation of two incompossible complete concepts. If
complete concept A, corresponding to possible individual a, includes a relational property
entailing aRb (i.e., if a exists, it is true that a bears R to b), and complete concept B,
corresponding to possible individual b, includes a relational property entailing ~(aRb), then it
would be logically inconsistent to suppose the existence of both a and b (Rescher 58). If, for
example, the complete concept of Queen Elizabeth II includes the property of being the mother
of Prince Charles (so it is true that, if she exists, she is the mother of the prince), then the Queen
could not exist in a world in which the complete concept of Prince Charles* (a counterpart of the
5
actual prince) includes the property of not being the son of Queen Elizabeth but rather the
illegitimate son of her sister. Such complete concepts would be incompossible.
For Rescher, a substance’s various relations to other substances are built in to its
complete concept by virtue of Leibniz’s universal expression thesis. As Leibniz presents it in
Discourse on Metaphysics, §9, the universal expression thesis states that ‘each individual
substance expresses the whole universe in its own way, and that all its events, together with all
their circ*mstances and the whole sequence of external things, are included in its notion’ (A
VI.iv 1541/AG 41). Rescher understands this to mean that there is a ‘conceptual linkage’
between compossible complete concepts, such that ‘no substance can—even in hypothesis—be
pried loose from its world-environment and transposed into some other possible world’ (49-50).
According to Rescher, Leibniz’s universal expression thesis implies not merely that it is logically
impossible for two incompossible substances to coexist, but that it is logically impossible for a
substance to exist apart from the unique set of substances with which it is compossible. Thus,
any world in which Queen Elizabeth II exists must be a world in which her son Prince Charles
exists. Furthermore, there is only one such world, for any substance is individuated through the
totality of its relations to the other members of its world. Change any one of these and
everything must change. Consequently, any substance is ‘world-bound’: it belongs to one and
only one possible world, which is a maximal set of compossible complete concepts.
Although Mates denies that relational properties are included in the complete concept of
a substance, because he sees this as inconsistent with Leibniz’s doctrine of the reducibility of
relations, he agrees with Rescher about the significance of the universal expression thesis (The
Philosophy of Leibniz 63, 219-20). Like Rescher, he takes the maximality constraint on possible
worlds and the world-boundedness of individuals to follow from the fact that the complete
6
concept of a substance expresses, and so ‘needs’, all of its worldmates. Given this, the
supposition of the existence of any one substance entails the existence of all its worldmates and
the non-existence of the members of different possible worlds.10
Although Mates and Rescher stress that compossibility is to be understood as a logical
relation among complete concepts, key features on their account stem from the role they assign
to the doctrine of universal expression. Their assertion of the maximality of possible worlds
does not follow from their definition of compossibility alone, but requires in addition the premise
that any member of a world ‘mirrors’ the rest of its world through its complete concept (Mates,
The Philosophy of Leibniz 76). Granting this assumption about the mutual relatedness of the
members of a world, and the way such relations are represented directly (Rescher) or indirectly
(Mates) by their complete concepts, it follows that the supposition of the existence of any one
member of a world entails the existence of all the other members of its world and the non-
existence of any substances that do not belong to that world.
The principal strength of the logical interpretation is that it provides Leibniz with a
cogent response to Spinoza: not all possibles are actual, because possibles organized into
different worlds on the basis of their mutual expression preclude each others’ existence. Given
the contents of their complete concepts, the supposition of the existence of any one substance
entails the existence of its worldmates and the nonexistence of the members of other possible
worlds.
However attractive this result, the logical interpretation faces several important
objections. There is, as we have noted, little direct evidence that Leibniz understands
compossibility in the way that Mates and Rescher do. The interpretation also is at odds with at
least one central Leibnizian doctrine: the ontological independence of substance. For Leibniz,
7
‘each substance is like a world apart, independent of all other things, except for God’ (A VI.iv
1550/AG 47; cf. A VI.iv 1541/AG 42; GP II 444/L 602). By this Leibniz seems to mean not
merely that each substance is immune to causal influence by other substances, but that each
substance could exist on its own apart from everything else but God. For Mates and Rescher,
however, every substance is so firmly moored to its worldmates that it is impossible for it to exist
without them. Aware of this problem, Mates concludes that Leibniz’s views on compossibility
require him to abandon the traditional doctrine of the ontological independence of substance
(The Philosophy of Leibniz 192, 221).
In addition to being in tension with the most natural reading of Leibniz’s world-apart
doctrine, the logical interpretation places a significant restriction on the scope of God’s power.
If it is logically impossible for a substance to exist without its worldmates, then not even God
can separate them, by actualizing one and not the others.11 However, various texts suggest that
Leibniz is open to just this possibility. That God actualized all the members of the actual world
rather than just some of them appears to be a matter of his wise and benevolent choice rather
than logical necessity.12
While the logical interpretation holds out the promise of a simple and rigorous definition
of compossibility, there is little evidence that Leibniz employs this definition in his philosophy
or that he regards compossibility merely as a logical relation among complete concepts. As
formulated by Mates and Rescher, the logical interpretation presupposes the holding of non-
logical relations of expression among substances, and it entails consequences that are at odds
with at least one central Leibnizian doctrine. For these reasons, we believe that the interpretation
fails as an adequate account of Leibniz’s position.
8
II
Advocates of the lawful interpretation deny that any logical contradiction is involved in
supposing the existence of incompossible substances. Absolutely speaking, God can actualize
any substance or set of substances. However, there are certain sets of substances that God would
not choose to create, because their actualization would fail to instantiate certain laws that God
wishes to uphold. For proponents of the lawful interpretation, two substances are compossible if
and only if they are suitably connected under some set of universal laws. As Ian Hacking puts
the point, ‘it is not logical inconsistency that prevents compossibility…. [Instead],
compossibility must be something like consistency under general laws of nature’ (193). The
laws in question are not metaphysically necessary, but rather contingent laws that conform to
God’s free decrees.
J. A. Cover and John O’Leary-Hawthorne defend such an account, motivated in part by
the apparent incompatibility of the logical interpretation with Leibniz’s belief in the ontological
independence of substance. They place a great deal of weight on the ‘world-apart’ passages,
which they take to show not merely that each substance is ontologically independent and
actualizable apart from other substances, but that God could constitute a world from any set of
substances, including the set of just a single substance. They thus deny the maximality
constraint on possible worlds, and allow one possible substance to belong to multiple possible
worlds, contrary to the thesis of world-bound individuals that many commentators have found in
Leibniz (135-40).
Since transworld identity would be impossible if a substance’s relations to a particular set
of substances were built into, or required by, its complete concept, Cover and O’Leary-
Hawthorne deny that the complete concept of a substance involves its relations to other
9
substances. Rather, the complete concept of a substance represents only its internal monadic
properties, namely, its perceptual states. Such perceptual states carry with them no commitment
to the existence of the things perceived (96-8). Once the complete concept of a substance is
unhooked from any connection to its worldmates in this way, it is possible to see how one
substance could belong to different worlds, while nevertheless retaining the same complete
concept, entailing the exact same history of monadic states.
Although Cover and O’Leary-Hawthorne deny that there are any metaphysically
necessary laws governing which substances can coexist in a world, they argue that compossible
substances in general instantiate hypothetically necessary laws—laws necessary on the
hypothesis of some possible divine decree. Such substances satisfy the condition they label
‘hypothetical compossibility’. Consistent with their account of substances as independent, mind-
like entities, they construe the relevant laws as ‘laws of expression’: laws describing the relation
between the contents of the perceptual states of different substances (98-9). They recognize,
however, that it is not enough to say simply that some lawful relationship must hold between the
perceptual states of any two compossible substances, for it is possible to find a lawful
relationship among any set of things, as Leibniz points out in Discourse on Metaphysics, §6 (A
VI.iv 1537-8/AG 39).13 Cover and O’Leary-Hawthorne thus conclude that some more specific
lawful correspondence must hold among the states of compossible substances, yet they have little
to say about what exactly this involves (134, 137; see also Wilson 129-31).
Cover and O’Leary-Hawthorne base their version of the lawful interpretation on the
notion of hypothetical compossibility. However, their account also reserves a place for the
broader notion of absolute or ‘per se compossibility’. As they see it, any two substances are per
se compossible, because they can be co-actualized by God (137; cf. 140). Cover and O’Leary-
10
Hawthorne are forced to say this in order to allow for the fact, which they take to be implied by
the world-apart doctrine, that any substance or set of substances can constitute a world. The
admission that all possible substances are per se compossible, though, leads to a serious
difficulty. If the set of all possible substances is a legitimate possible world, then how does
Leibniz’s doctrine of compossibility overcome the threat of Spinozism that it was designed to
avoid? Cover and O’Leary-Hawthorne have two responses open to them. First, since there is no
maximality constraint on possible worlds, the possible world consisting of all possible
substances is not the only possible world; there are alternatives, among which God can choose.
Second, although all possible substances are absolutely compossible, they are not hypothetically
compossible, since the actualization of all possible substances would not instantiate the sorts of
laws of expression that God prefers. There would not be the appropriate level of agreement
among the representational states of all possible substances for God to want to actualize them,
although he could make a world from them in principle (135-7).
Neither of these responses gets around the main difficulty. The compossibility relation is
introduced by Leibniz to explain why God does not actualize all possible substances. Rescher
and Mates have a ready explanation: there is no such possible world, because certain possibles
logically exclude one another. Cover and O’Leary-Hawthorne’s account of compossibility, by
contrast, does not so much explain God’s choice as presuppose it. Furthermore, it is not obvious
that God would not choose to actualize a world consisting of all possible substances if such a
world were indeed possible. Cover and O’Leary-Hawthorne assume that God’s preference for
harmony would trump his interest in diversity and plenitude when faced with a world containing
all possible substances. Yet there is evidence that Leibniz’s God is disposed to actualize as
many possibles as he can, consistent with those possibles forming a single world.14
11
The foregoing is an objection that applies to all versions of the lawful interpretation, not
just Cover and O’Leary-Hawthorne’s. If the compossibility relation is to serve as the basis of a
satisfying response to Spinoza, it cannot be ‘up to God’ which possible substances are
compossible with one another. If there were no logical or metaphysical obstacles to God
creating a world from all possible substances, then it is natural to assume that God would create
such a world.15 It is not a lack of will or desire that explains why God fails to do so, but rather
the incompossibility of certain substances. For proponents of the lawful reading, however, facts
about which substances are compossible (and thus which substances can belong to the same
possible world) are ‘up to God’. On their view, God freely decrees certain laws for a world, and
the compossibility or incompossibility of any set of substances is a function of whether or not
they instantiate the relevant laws. In effect, proponents of the lawful reading simply assume that
God would not choose to actualize a world consisting of all possible substances, rather than
explain why he could not do so.
The version of the lawful interpretation defended by Cover and O’Leary-Hawthorne
gives due weight to the ontological independence of substance and to the role played by laws in
uniting substances in a world. As it stands, however, it lacks the resources needed to ground a
suitably restrictive notion of compossibility. Taking ‘law’ in its most general sense, any
collection of possible substances can be conceived as compossible for Leibniz. In order to meet
this objection, proponents of the lawful interpretation may propose restricting the class of world-
ordering laws to some proper subset of laws (‘laws of expression’, ‘laws of nature’). Such
attempts, however, risk appearing arbitrary unless some further account is offered of the role
played by laws in the constitution of a world.16
12
III
In addition to the logical and lawful interpretations, there are hybrid readings that attempt to
accommodate the notion of lawfulness within a logical account of compossibility. Margaret
Wilson defends such a reading. While she believes the logical interpretation is ‘basically
correct’ (121), she points to several passages that demonstrate the importance of the notion of
lawfulness for the compossibility relation. One such passage occurs in Leibniz’s correspondence
with Arnauld:
There were an infinity of possible ways of creating the world, according to the
different designs which God might form, and each possible world depends upon
certain principal designs or ends of God proper to itself, i.e. certain free primitive
decrees (conceived sub ratione possibilitatis), or laws of the general order of this
possible universe, to which they belong, and whose notion they determine, as well
as the notions of all the individual substances which must belong to this same
Universe. (GP II 51; quoted in Wilson 128)
Wilson takes this passage to confirm the role played by laws in uniting substances in a world, a
role stressed by the lawful reading. At the same time, she believes that the lawful reading is
inadequate because it fails to make room for the further claim that incompossible substances are
logically inconsistent. This claim is central to the logical interpretation.
Wilson presents two versions of a logical reading that incorporates considerations of
lawfulness.17 She ascribes the first to Bertrand Russell. On Russell’s account, two substances
are compossible if and only if they stand in lawful relations with one another. Russell denies that
the complete concepts of incompossible substances directly contradict one another. In this he
sides with the lawful reading. Nevertheless, Russell agrees with the logical reading that it is
13
metaphysically impossible for two incompossible substances to coexist. This is because, by
Russell’s lights, the supposition that coexisting substances fail to instantiate lawful relations is
incompatible with the Principle of Sufficient Reason (PSR). Because Russell takes PSR to entail
that God cannot actualize a given set of substances without a sufficient reason, and because he
believes that only sets of substances that instantiate laws have a possible sufficient reason for
their existence, he concludes that incompossible substances cannot be created (Russell 66-7).
Russell’s reading is vulnerable to an objection we considered above: the requirement that
compossible substances stand in lawful relations with one another is vacuous, because it is
possible to find a lawful relation among any set of things. Wilson suggests a reply to this
objection. She proposes that the laws relevant to compossibility are ‘fairly simple lawful
generalities’, e.g. E = mc2. Since such laws place strict limits on which substances can belong to
a common world, Russell can evade the vacuity objection (Wilson 130-1). As we have argued,
however, restrictions of this sort risk appearing arbitrary unless more is said about the role
played by laws of nature in the constitution of a possible world.
The second hybrid reading that Wilson describes, which she herself favors, builds on
Russell’s account, while avoiding its reliance on the Principle of Sufficient Reason. Wilson
argues that the complete concept of a substance contains the laws of nature of the world to which
it belongs. She proposes that we regard these laws as possible facts. If two complete concepts
contain incompatible laws of nature – e.g. one contains E = mc2, while the other contains E =
2mc – then they are logically inconsistent, because they imply contradictory facts. Although
Russell and Wilson both invoke considerations of lawfulness to explain how the supposition of
the existence of incompossible substances entails a contradiction, they differ on a crucial point.
While Russell denies that the complete concepts of incompossible substances are logically
14
inconsistent, Wilson claims that they are by virtue of containing incompatible laws of nature
(Wilson 131-2).
Wilson’s hybrid reading remains close to the logical interpretation of Mates and Rescher,
and it faces the same challenges that we noted regarding their account. Her discussion
nevertheless brings to light two plausible constraints on any adequate explanation of
compossibility. First, any such explanation must incorporate the idea that compossible
substances are united in a world by virtue of their mutual relation, or ‘connection’, under general
laws of nature. We will call this the Connection Condition. The Connection Condition is
explicitly acknowledged by the lawful interpretation and by Wilson’s hybrid reading. Second,
any adequate account of compossibility must accept that the complete concept of a substance
includes its relations to its worldmates in a way that sets limits on which sets of complete
concepts God can actualize as a world. We will call this the Inclusion Condition. The Inclusion
Condition is supported by Leibniz’s statement that the complete concept of a substance includes
both its intrinsic and extrinsic denominations (GP II 56/M 63-4), and by the role he assigns to the
compossibility relation in formulating a response to Spinoza. God is limited to creating one from
among an infinity of possible worlds, because, given the content of their complete concepts, the
existence of certain possibles precludes the existence of others as members of the same world.
The necessity of the Inclusion Condition is stressed by proponents of the logical interpretation
and by Wilson.
We regard the Connection and Inclusion Conditions as integral to Leibniz’s account of
compossibility and trace the weaknesses of previous interpretations to their failure to adequately
meet both of these conditions. Standard versions of the logical interpretation neglect the
Connection Condition and construe the Inclusion Condition too strongly, with the result that God
15
is unable to actualize the complete concept of a substance without actualizing those of its
worldmates. This, we have seen, is at odds with Leibniz’s ‘world-apart’ doctrine and his
understanding of divine omnipotence. By contrast, standard versions of the lawful interpretation
reject the Inclusion Condition, with the result that they are forced to admit as possible a world
containing all possible individuals—thus robbing Leibniz of a cogent response to Spinoza.
Wilson’s hybrid reading goes beyond the logical and the lawful interpretations insofar as it aims
to satisfy both conditions. Like the logical interpretation, however, Wilson’s hybrid reading
construes the Inclusion Condition too strongly, so that God is incapable of actualizing a
substance apart from its worldmates.
On the face of it, it may seem doubtful that any interpretation can meet the requirements
of both the Connection and Inclusion Conditions. In making the lawful relation of possibles
sufficient for their compossibility, the lawful interpretation undercuts the force of Leibniz’s
response to Spinoza. On the other hand, by rendering those relations part of the complete
concept of any substance, the logical interpretation threatens Leibniz’s view of the ontological
independence of substance and his commitment to divine omnipotence. Can any interpretation
accommodate both of these requirements while remaining consistent with the central doctrines of
Leibniz’s metaphysics? We believe that the cosmological interpretation developed in the next
section can.
IV
On the logical interpretation, Leibniz explains membership in a world in terms of compossibility.
Two or more substances are members of the same world by virtue of the compossibility of their
complete concepts. In our view, this reverses the correct order of explanation. The proper way
16
to frame the issue of compossibility, we believe, is to begin with the notion of a world, as an
abstract relational structure according to which God conceives of possibilities of existence. On
the cosmological interpretation we defend, two or more substances are compossible by virtue of
the fact that God can conceive of them as belonging to the same world; by contrast, substances
are incompossible because God cannot conceive of them as belonging to the same world. The
emphasis we place on what God can and cannot conceive signals our alignment with the logical
reading on one key point. Since for Leibniz the set of things that God cannot conceive is
coextensive with the set of things God cannot do, the cosmological interpretation is intended to
provide an account of compossibility that is sufficiently strong to ground a response to Spinoza:
not all possibles are compossible, because God cannot conceive of all possibles as belonging to
the same world. This is in contrast to the conclusion of the lawful interpretation, which barring
arbitrary restrictions, is limited to explaining compossibility in terms of what God would not
choose to do, as opposed to what he cannot do.
The plausibility of the cosmological interpretation hinges on our being able to make good
on the thesis that compossibility is best explained in terms of whether God can conceive of
certain substances as belonging to the same world. This requires that we substantiate two main
claims: first, that the notion of a world is conceptually prior to the notion of compossibility;
second, that God’s knowledge of any possible substance is structured such that, in knowing it,
God conceives of the substance as it would exist in a world related to other substances, while at
the same time conceiving of it in such a way that he could create that substance by itself
separated from its world, though not as a member of some other world. Our defense of these two
theses will demonstrate how the cosmological interpretation satisfies the Connection and
Inclusion Conditions. We begin with the Connection Condition, which is closely associated with
17
the claim that for Leibniz the notion of a world is conceptually prior to the notion of
compossibility.18
The Connection Condition
The primary reason for taking the relation of connection as foundational to Leibniz’s conception
of a possible world is that it is how he presents his own position. In Theodicy, §9 he writes: ‘For
it must be known that all things are connected [tout est lié] in each one of the possible worlds:
the universe, whatever it may be, is all of one piece, like an ocean: the least movement extends
its effect there to any distance whatsoever, even though this effect becomes less perceptible in
proportion to the distance’ (GP VI 107/H 128; Leibniz’s emphasis). By the ‘connection’ of all
things, Leibniz means a mutual dependence among the states of substances, such that a change in
any one substance is reflected in a corresponding change in every other. Although he denies that
substances exert any real, or ‘metaphysical’, influence on each other, he holds that the members
of a world condition each others’ existence, in accordance with contingent laws of nature.19 In
the same section of the Theodicy, Leibniz further stresses that the ideal dependence of substances
on one another is conceived by God prior to the decision to create those substances (or any
others): ‘God has ordered all things beforehand once for all, having foreseen prayers, good and
bad actions, and all the rest; and each thing as an idea [idealement] has contributed, before its
existence, to the resolution that has been made upon the existence of all things; so that nothing
can be changed in the universe’ (GP VI 107-8/H 128-9; Leibniz’s emphasis). Thus, it is integral
to God’s knowledge of a set of substances as a possible world that they are conceived as
conditioning each others’ existence in a lawful manner.
18
This aspect of Leibniz’s position is endorsed by the lawful interpretation, which goes on
to claim that compossibility is fully explained by the laws that God decrees to exist among the
states of substances. We reject this inference, both because (as we have argued) it supports too
weak a conception of compossibility and because it does not adequately reflect Leibniz’s
understanding of the relational structure of a world. In Theodicy, §8 Leibniz makes it clear that,
in addition to their connection, the substances that make up a world must be united within a
common spatiotemporal order:
I call a world the entire series and entire collection of all existing things, lest it be
said that several worlds could have existed at different times and different places.
For they must be reckoned all together as one world or, if you will, as one
universe. And even though one should fill all times and all places, it still remains
true that one could have filled them in infinite ways, and that there is an infinity
of possible worlds, from among which God must have chosen the best, since he
does nothing without acting in accordance with supreme reason. (GP VI 107/H
128)
Although Leibniz begins by speaking of the actual world (the ‘collection of all existing things’),
he goes on to affirm that there is an infinity of possible worlds, which are distinguished (in part)
by the ways in which things are spatially and temporally ordered within them. The implication is
that for a set of things to count as ‘one world’ each of them must be spatiotemporally ordered
with respect to every other member of the world, and nothing that is not a member of the world
can have a spatiotemporal relation with respect to anything that is a member of the world.
19
We find explicit confirmation of this idea in many texts. For Leibniz, space and time—or
as he also designates them, the order of coexistence and the order of succession—represent an
abstract structure within which possible things can be conceived as belonging to the same world:
Space and time taken together constitute the order of possibilities of one entire
universe, so that these orders—space and time, that is—relate not only to what
actually is but also to anything that could be put in its place. (GP IV 568/L 583)
The description of space and time as ‘orders of possibilities of one entire universe’, requires
comment. As is well known, Leibniz rejects the Newtonian view of space and time as real or
absolute entities within which bodies are located. In his exchange with Samuel Clarke, he also
insists that space and time cannot be reduced to any set of actual relations among things.
Instead, space and time are ideal orders that specify the possible ways that things can be
conceived to coexist or to succeed one another. To Clarke, he writes: ‘I hold space to be an
order of coexistences, just as time is an order of successions. For space denotes, in terms of
possibility, an order of things that exist at the same time, considered as existing together, without
entering into their particular manners of existing’ (III, §4; GP VII 363/AG 324, emphasis
added).20
The last qualification is critical to our defense of the cosmological interpretation. For
Leibniz, space and time, the orders of coexistence and succession, have a broad metaphysical
significance. They are relevant not just to physical theory, defining possible modes of existing
of bodies, but to any conception of a world, including the most basic notion of possible worlds as
collections of compossible substances. According to the cosmological interpretation, substances
can be conceived as belonging to the same world, within which they are lawfully connected, only
if they are related according to the orders of space and time. If individuals cannot be understood
20
as related in this way, then they cannot be members of the same world—a point Leibniz
illustrates with the example of fictional worlds such as that of Thomas More’s Utopia. While the
worlds described in works of fiction are possible in themselves, they could not be part of our
world because their inhabitants lack any spatiotemporal relation to us.21
A full account of the cosmological interpretation would require discussion of how
Leibnizian substances (including soul-like monads) can be understood as spatiotemporally
ordered.22 Here we defer elaboration of the details of Leibniz’s theory and focus on its
application to the question of compossibility. The basic insight of Leibniz’s position is that
substances can be conceived as connected in a world, only if they are related with respect to a
common spatiotemporal order. To the extent that substances are connected, they condition each
others’ existence. This means that the states of some substances are understood to explain the
states of other substances, in accordance with the laws of their world. The general form of such
laws will involve appeal to how dependence relations among the states of substances vary with
respect to their relative spatiotemporal positions.
In Leibniz’s view, all and only those substances are compossible that are conceived by
God as related within the spatiotemporal and causal structure of a world. They must, in other
words, be substances that condition each others’ existence, relative to a common order of
coexistence and succession. Whatever things cannot be conceived by God in this way are
incompossible. The addition of the constraint of a common order of coexistence and succession
blocks in a natural way the threat of a single world consisting of all possible substances. Leibniz
reasonably assumes that there are possible things that are not spatiotemporally related to us. If
spatiotemporal relatedness is necessary for membership in a world, then not all possibles are
members of one world. Hence, in conceiving of the laws governing the connection of
21
substances, we may limit ourselves to those that hold of spatiotemporally ordered worlds. In
response to this restriction, it might be objected that limiting the notion of a possible world to
sets of substances that have a well-defined spatiotemporal order is no less arbitrary than limiting
it to sets of substances that are related by a privileged class of natural laws. The difference in our
view is that there is evidence that this is the path by which Leibniz developed his answer to
Spinoza: not all possibles are compossible, because not all possibles can be arranged within a
common order of coexistence and succession. To the extent that the latter defines the relational
structure of a world, and God chooses to create a single world, God is limited to creating some,
but not all, possibles.23
According to the cosmological interpretation, in conceiving of an individual substance,
God conceives of that substance as it would be related, spatiotemporally and causally, to the
other members of its world. All of this information is contained in the complete concept of a
substance. Hence in deliberating about whether to create that substance, God necessarily
deliberates about whether to create the world of which it is a part. On the assumption that God’s
aim is to create a single world (‘the best of all possible worlds’), God is bound to create all those
substances to which a given substance is spatiotemporally and causally related and no substances
to which it is not so related. The former follows from the fact that any substance is partly
defined through the relations it bears to the other substances in its world; thus, to create a world
in which that substance exists, God must create all the substances to which it is related.24 At the
same time, in creating a single world, God is limited to creating just the individuals to which that
substance stands in the requisite spatiotemporal and causal relations. The thought that other
substances besides these might be created is blocked by the requirement that to be members of
the same world substances must be united by a common spatiotemporal and causal order.
22
The Inclusion Condition
We have argued that the principal problem with the logical interpretation is that it construes the
Inclusion Condition too strongly, so that not even God can pry a substance apart from its
worldmates, by actualizing one and not the others. Defenders of the logical interpretation face
this problem because they take the complete concept of a substance to include the actual
relations it would stand in to its worldmates were it to exist. Given this understanding of how a
substance’s relations are contained in its complete concept, God cannot actualize any substance
without actualizing its worldmates, since the existence of the latter is entailed by the existence of
the former.
On the cosmological reading, every substance is conceived by God as belonging to a
single possible world. The content of God’s knowledge of each substance is such that were he to
create it as part of a world, he would be bound to create all the members of that world and no
substances that are not members of that world. Since we follow the logical interpretation this far,
it might thought that the cosmological interpretation also must be committed to the conclusion
that God cannot create a substance without creating its worldmates. Since we have criticized the
logical interpretation on this count, arguing that it is inconsistent with Leibniz’s understanding of
divine omnipotence and the ‘world-apart’ doctrine, this would be an unfortunate result. In fact,
we believe that this conclusion does not follow on our reading, provided that one is careful to
distinguish what God can do absolutely and what God can do in meeting the objective of
actualizing a world.
In order to enforce this distinction, it is necessary to draw a sharp line between the way in
which two different sorts of properties—those designated by intrinsic and extrinsic
23
denominations—are contained in the complete concept of a possible substance. In conceiving of
a substance as possible, God conceives of both the internal states by which it would be modified
were it to exist and the ways in which its states would be related, spatiotemporally and causally,
to the states of the other members of its world, were that world to exist. God’s conception of the
internal states of any substance presupposes nothing about the internal states of its worldmates;
he conceives of the substance as ‘a world apart’. By contrast, God’s conception of a substance’s
extrinsic denominations necessarily involves an idea of how it would be related to the other
members of its world, in particular, the ways in which their states would be ‘connected’.
Substances are connected in a world in accordance with contingent, causal laws that God freely
decrees in choosing to actualize that world. Thus, in conceiving of the extrinsic denominations
of a substance, God conceives of possible free decrees he would exercise in bringing its world
into existence.25
The content of a substance’s complete concept tracks its membership in a world by
representing the relational properties that the substance would have were its world to exist. If
God chooses to create that world and to enact the free decrees associated with it, then the
substance exists with those relational properties. Yet nothing in Leibniz’s complete concept
theory precludes God from actualizing a substance apart from its worldmates and hence without
the relational properties specified by its complete concept. To do so, God has only to decide not
to enact the free decrees associated with the creation of that world and instead to enact different
free decrees associated with the existence of a solitary substance. A substance created under this
scenario (‘separated’ from its world) would have the same complete concept, specifying modally
the relations it would stand in were its world to exist, but the divine free decrees on which those
relations depend would remain merely possible.
24
On this way of understanding Leibniz’s position, the content of a substance’s complete
concept is identified with God’s prevolitional knowledge of the substance, that is, the knowledge
God has of its properties independently of his knowledge of his own actual free decrees.26 God
knows a possible substance as an individual in knowing the intrinsic properties (e.g. perceptual
states) it would have were it to exist and the relational properties it would have were the other
members of its world to exist. Given this account of divine knowledge, we suggest, God can
create the same individual under difference circ*mstances—with or without the other members
of its world. The difference in God’s knowledge of the two cases is explained by his knowledge
of the different free decrees that would be exercised in them, not by the content of the relevant
complete concept(s). If God chooses to create a world, specified in terms of the free decrees that
define its contingent causal structure, he is committed to creating all and only those substances
that comprise that world. But God could decide not to create a world, choosing instead to create
one or more separated substances, which lacked the unity of a world. In this case God would
actualize the individual substance without actualizing the free decrees contained in its complete
concept. Thus, God would create the substance, without creating it as part of a world. As
Leibniz emphasizes, there is no reason to think God would do this. Nevertheless, it remains
something that God could do.
V
We have argued that any adequate account of Leibniz’s notion of compossibility must
accommodate both the Connection and Inclusion Conditions. The cosmological interpetation
meets these conditions in a way that is illuminating of Leibniz’s response to Spinoza’s
necessitarianism and consistent with the foundational commitments of his metaphysics. On our
25
account, compossibility is to be understood as the possibility of existence within a common
world, where a world is defined in terms of a unified spatiotemporal and causal order. The
abstract relational structure that defines a possible world implies for Leibniz that not all possibles
are compossible, for all possibles are not related within a single order of coexistence and
succession. Thus, on the assumption that God aims to create a unique world, he is limited to
creating some but not all possible substances.
In conceiving of any possible substance, God conceives of that substance as it would be
related to the other members of its world. Thus, the decision to create that substance as part of a
world entails a decision to create all the other members of its world and no substances that are
not members of its world. We have argued, however, that nothing in Leibniz’s philosophy
implies that God could not also create that substance by itself, separated from any world. This
would require only that God choose not to actualize the possible free decrees that are definitive
of the causal order of that world and choose instead to create the substance outside the structure
of a world. The assumptions Leibniz makes about God’s will—in particular, that God acts only
for the best—leave us with no reason to believe that God would choose to act in this way.
Nevertheless, it is a strength of our account that it makes sense of how, in Leibniz’s view, this is
something God could do.27
26
Works Cited
Adams, Robert. Leibniz: Determinist, Theist, Idealist. New York: Oxford UP, 1994.
Anfray, Jean-Pascal. ‘God’s Decrees and Middle Knowledge: Leibniz and the Jesuits’.
American Catholic Philosophical Quarterly 76 (2002): 647-70.
Broad, C. D. Leibniz: An Introduction. Cambridge: Cambridge UP, 1975.
Brown, Gregory. ‘Compossibility, Harmony, and Perfection in Leibniz’. Philosophical Review
96 (1987): 172-203.
Cover, J.A. and John O’Leary-Hawthorne. Substance and Individuation in Leibniz. Cambridge:
Cambridge UP, 1999.
D’Agnostino, F. B. ‘Leibniz on Compossibility and Relational Predicates’. Leibniz:
Metaphysics and Philosophy of Science. Ed. Roger Woolhouse. Oxford: Oxford UP,
1981. 89-103.
Di Bella, Stefano. The Science of the Individual: Leibniz’s Ontology of Individual Substance.
Dordrecht: Springer, 2005.
Hacking, Ian. ‘A Leibnizian Theory of Truth’. Leibniz: Critical and Interpretative Essays. Ed.
Michael Hooker. Minneapolis: University of Minnesota Press, 1982. 185-95.
Hintikka, Jaakko. ‘Leibniz on Plenitude, Relations, and “The Reign of Law”’. Leibniz: A
Collection of Critical Essays. Ed. Harry G. Frankfurt. Garden City, NY: Doubleday,
1972. 335-64.
Koistinen, Olli and Arto Repo. ‘Compossibility and Being in the Same World in Leibniz’s
Metaphysics’. Studia Leibnitiana 31 (1999): 196-214.
27
Kulstad, Mark. ‘Leibniz, Spinoza and Tschirnhaus: Multiple Worlds, Possible Worlds’. The
Young Leibniz and his Philosophy. Ed. Stuart Brown. Dordrecht: Kluwer, 1999. 243-
62.
Leibniz, Gottfried Wilhelm. De Summa Rerum: Metaphysical Papers, 1675-1676. Tr. G. H. R.
Parkinson. New Haven: Yale UP, 1992.
-------. Die Philosophische Schriften von Gottfried Wilhelm Leibniz. Ed. C. I. Gerhardt. Berlin:
Weidmann, 1875-90; repr. Hildesheim: Olms, 1960.
-------. The Labyrinth of the Continuum: Writings on the Continuum Problem, 1672-1686. Ed.
and tr. R. T. W. Arthur. New Haven: Yale UP, 2001.
-------. The Leibniz-Arnauld Correspondence. Ed. and tr. H. T. Mason. Manchester:
Manchester UP, 1967.
-------. New Essays on Human Understanding. Ed. and tr. P. Remnant and J. Bennett.
Cambridge: Cambridge University Press, 1981.
-------. Opuscles et fragments inédits de Leibniz. Ed. L. Couturat. Paris: Alcan, 1903.
-------. Philosophical Essays. Ed. and tr. R. Ariew and D. Garber. Indianapolis: Hackett, 1989.
-------. Philosophical Papers and Letters. Ed. and tr. L. E. Loemker, 2nd ed. Dordrecht: Reidel,
1969.
-------. Sämtliche Schriften und Briefe. Ed. Deutsche Akademie der Wissenschaften. Darmstadt
and Berlin: Akademie-Verlag, 1923ff.
-------. Textes inédits d’après les manuscrits de la Bibliothèque provinciale de Hanovre. Ed.
Gaston Grua. Paris: PUF, 1948.
-------. Theodicy. Tr. E. M. Huggard. LaSalle: Open Court, 1985.
28
Mates, Benson. ‘Leibniz on Possible Worlds’. Leibniz: A Collection of Critical Essays. Ed.
Harry G. Frankfurt. Garden City, NY: Doubleday, 1972. 335-64.
-------. The Philosophy of Leibniz: Metaphysics and Philosophy of Language. New York:
Oxford UP, 1986.
Rescher, Nicholas. Leibniz: An Introduction to His Philosophy. Lanham, MD: University Press
of America, 1979.
Rutherford, Donald. Leibniz and the Rational Order of Nature. New York: Cambridge UP,
1995.
Russell, Bertrand. A Critical Exposition of the Philosophy of Leibniz, 2nd ed. London: George
Allen and Unwin, 1937.
Sleigh, R. C., Jr. Leibniz and Arnauld: A Commentary on Their Correspondence/ New Haven:
Yale UP, 1990.
Spinoza, Baruch. The Ethics and Other Works. Ed. and trans. Edwin Curley. Princeton:
Princeton UP, 1994.
Wilson, Margaret. ‘Compossibility and Law’. Causation in Early Modern Philosophy. Ed.
Steven Nadler. University Park, PA: Pennsylvania State UP, 1993. 119-33.
29
Notes 1 Ethics, Part I, Proposition 29.
2 Concern about Spinoza’s necessitarianism permeates Leibniz’s thought until the end of
his life. See the notes composed shortly after their meeting in November 1676 (A VI.iii
582/DSR 105), as well as GP II 55-6/M 62-3 and Theodicy, §§173-174 (GP VI 217-8/H 234-6).
Editions of Leibniz’s writings are cited according to the following abbreviations: A = Gottfried
Wilhelm Leibniz: Sämtliche Schriften und Briefe, cited by series, volume, and page; AG =
Philosophical Essays; C = Opuscles et fragments inédits; DSR = De Summa Rerum:
Metaphysical Papers, 1675-1676; GP = Die Philosophische Schriften von Gottfried Wilhelm
Leibniz, cited by volume and page; Grua = Textes inédits; H = Theodicy; L = Philosophical
Papers and Letters; LOC = The Labyrinth of the Continuum: Writings on the Continuum
Problem, 1672-1686; M = The Leibniz-Arnauld Correspondence. Where a translation is cited,
we have relied on it, though we have sometimes taken the liberty of modifying it slightly; where
none is cited, the translation is our own.
3 We say ‘at a minimum’, because the problem of contingency has several dimensions and
elicits a variety of responses from Leibniz. For a discussion of these, see Adams 9-110.
4 Wilson (119) notes these two explanatory functions of Leibniz’s notion of
compossibility. The claim that the compossibility relation ‘partitions’ possible substances into
worlds is defended by Mates (‘Leibniz on Possible Worlds’ 341; The Philosophy of Leibniz 77).
Talk of ‘possible substances’ may be read as shorthand for the complete concepts of possibles, as
represented in the divine mind. To Arnauld, Leibniz writes: ‘purely possible substances… have
no other reality than that which they have in the divine understanding and in the active power of
30
God. However, you can see from that, Sir, that one is obliged to have recourse to divine
knowledge and power in order to explain them properly’ (GP II 54/M 61).
5 Mates, Rescher, Hintikka, and D’Agostino defend versions of the logical interpretation.
6 Hacking, and Cover and O’Leary-Hawthorne, defend versions of the lawful
interpretation. Russell is also commonly associated with the lawful reading; however, Wilson
(127-31) argues that Russell is better understood as offering a logical interpretation that
incorporates elements of the lawful account. Wilson (120-1) proposes the contrast between
‘logical’ and ‘lawful’ construals of compossibility as an alternative to the distinction between
‘analytic’ and ‘synthetic’ readings introduced by D’Agostino. Wilson’s terminology is adopted
by Cover and O’Leary-Hawthorne (132).
7 There is controversy about which properties of a substance are included in its complete
concept and which follow from the concept, taken together with the concepts of other individuals
and the laws of its world. Mates, and Cover and O’Leary-Hawthorne, maintain that a complete
concept includes only the ‘intrinsic denominations’ of a substance, not its ‘extrinsic
denominations’, or relational properties; Rescher maintains that a complete concept includes all
of a substance’s properties, including its relational ones.
8 See also D’Agostino, and Rescher 55 for this point.
9 The passage most often cited on its behalf is tentatively dated ca. 1687-96: ‘The
compossible is that which, with another, does not imply a contradiction’ (A VI.iv 867/Grua 325).
Mates (The Philosophy of Leibniz 75, n. 36) offers a number of other texts in support of the
logical interpretation. None is conclusive. Of greatest interest is a note, cited by its Bodemann
catalogue number (LH IV 7B 3, Bl. 17v), which has now appeared in the Akademie edition of
Leibniz’s writings: ‘If, given the proposition “A exists”, the proposition “B does not exist”
31
follows, then A and B are incompatible (incomponibile)’ (A VI.iv 401). Di Bella (240-1) makes
a convincing case that the notion of ‘incompatibility’ introduced here should not be identified
with logical contradiction.
10 Mates distinguishes between a concept ‘needing’ and ‘including’ another. One concept
includes another if it is impossible for something to fall under the first without also falling under
the second, or if analysis of the first reveals the second to be contained it in. The ‘needing’
relation is much weaker. According to Mates, ‘The concept of A needs or leads to the concept of
B if and only if some part of the first expresses the second’ (The Philosophy of Leibniz 220). As
his examples make clear, two complete concepts express each other just in case there would be a
relation of agreement, or ‘mirroring’, between the perceptions of their corresponding substances
(The Philosophy of Leibniz 76, 78).
11 Addressing this point, Rescher writes: ‘It is a fundamental tenet of Leibniz’s philosophy
that even omnipotence cannot accomplish the impossible’ (54). We agree: the question is
whether or not incompossibility is to be analyzed in terms of logical impossibility.
12 To Des Bosses’s objection that ‘God cannot have created any of these monads which now
exist without having constituted all the rest’, Leibniz responds: ‘He can do it absolutely; he
cannot do it hypothetically, because he has decreed that all things should function most wisely
and harmoniously’ (GP II 496/L 611). See also GP IV 530, and Discourse on Metaphysics, §14:
‘This [sequence of thoughts and perceptions] would never fail, and it would happen to me
regardless, even if everything outside me were destroyed, provided there remained only God and
me’ (A VI.iv 1550/AG 47).
13 For a development of this point, see Brown 179.
32
14 See Theodicy, §201 (GP VI 236/H 252-53); GP VII 289; GP VII 304/AG 151; A VI.iii
472/LOC 45; A VI.iv 1395/LOC 239.
15 For an argument that this should be Leibniz’s position, see Broad 162.
16 Although it is tangential to our criticisms of the lawful interpretation, Cover and
O’Leary-Hawthorne’s allowance of transworld identity also raises problems. As they admit (93-
4), several prominent texts suggest that Leibniz takes substances to be world-bound. However,
they assert that (1) transworld identity is implied by the world-apart doctrine, and (2) transworld
identity is not ruled out by Leibniz’s complete concept theory, because the complete concept of a
substance contains only its intrinsic properties and not its relational ones. With regard to (1), it is
not obvious that transworld identity is implied by the world-apart doctrine. Leibniz says that any
substance is like, or as it were, a world apart, not that any single substance could literally
constitute a world. With regard to (2), there is good evidence that Leibniz believed that a
substance’s relations to other substances are included in its complete concept, and that he took
this inclusion to be implied by his theory of truth (see GP II 37/M 40; GP II 49/M 55; and
especially, GP II 56/M 63-4). Such evidence supports the conclusion that, while causally
autonomous, substances are ‘world-bound’ in at least this sense: insofar as God conceives of
them as belonging to a world, that world is unique.
17 A different hybrid interpretation is developed by Brown.
18 Koistinen and Repo advance a reading which like ours makes central the idea of a world.
In doing so, they acknowledge (207, n. 34) the precedent of the view defended in Rutherford. In
other respects, however, our accounts differ significantly. Koistinen and Repo argue (207-11)
that the connection that unites substances in a world and renders them compossible is a universal
harmony among their states. Sleigh (170-82) offers reasons for doubting whether universal
33
harmony is a feature of every possible world for Leibniz. Over and above this, we question
whether the notion of harmony is sufficiently precise or basic to ground an explanation of
compossibility.
19 According to Leibniz, the connection among substances depends upon God’s free decrees
(GP II 56/M 63), which are the basis of the ‘laws of the general order’ of any possible world (GP
II 51/M 57). On the mutual dependence of substances, see Primary Truths: ‘Every individual
created substance exerts physical action and passion on all the others. From a change made in
one, some corresponding change follows in all the others, since the denomination is changed….
Strictly speaking, one can say that no created substance exerts a metaphysical action or influx on
any other thing…. What we call causes are only concurrent requisites, in metaphysical rigor’ (A
VI.iv 1646-7/AG 33); and On Nature Itself: ‘In another place I shall give a better account of
what can be said about the transeunt actions of created things. Indeed, elsewhere I have already
explained a part of it, namely, that the correspondence [commercium] among substances or
monads arises not from an influx, but through an agreement derived from divine preformation’
(GP IV 510/AG 161).
20 See also New Essays, II.xiv.24, 26: ‘time and space pertain as much to possibles as to
existents…. As I have just said, time and space indicate possibilities beyond any that might be
supposed to be actual. Time and space are of the nature of eternal truths, which equally concern
the possible and the actual’ (A VI.vi 153-4).
21 ‘I do not agree that “in order to know if the romance of Astrea is possible, it is necessary
to know its connection with the rest of the universe”. It would be necessary to know this if it is
compossible with the universe, and, consequently, if this romance has taken place, is taking
34
place, or will take place in some corner of the world, for certainly there would be no place for it
without such a connection’ (GP III 572/L 661). See also A VI.iv 1653-4/L 263; GP II 181.
22 For more on this, see Rutherford 188-97.
23 See A VI.iii 511/DSR 65; A VI.iv 1397. The supposition that God aims to create a
unique world is critical to Leibniz’s position. Although all possibles cannot be created by God
as a single world, it might be objected that nothing prevents God from actualizing many
spatiotemporally disjoint worlds (Kostinen and Repo 213-4). Leibniz explores this issue in texts
composed during his most intense engagement with Spinoza’s philosophy (1676-77). Whether
or not he arrives at a compelling reason for rejecting this scenario, he is firm that God chooses to
create only one from among an infinity of possible worlds. For a discussion of the relevant texts,
see Kulstad.
24 The last assumption might be questioned. In particular, it might be wondered why there
could not be a possible world w2 whose members were qualitatively identical (by virtue of
possessing the same intrinsic denominations) to a proper subset of the members of a maximal
world w1. This scenario is blocked by Leibniz’s ‘no purely extrinsic denominations’ thesis,
which itself is a corollary of the Connection Condition. According to the no purely extrinsic
denominations thesis, any change in a relational property of a substance entails some change in
an intrinsic property of it, by virtue of the way that substances condition each others’ existence in
a world (C 8, 520-1; A VI.vi 227). In the case of the imagined scenario, the members of w2
would not be conditioned by (all) the same substances as their counterparts in the maximal world
w1. Hence, their intrinsic denominations (e.g. their perceptual states) would have to be different
from those of their counterparts in w1. For further discussion of this point, see Rutherford 145.
35
25 Leibniz develops this point in his letter to Arnauld of 4/14 July 1686 (GP II 49-51/M 54-
7).
26 For a defense of this reading, see Anfray.
27 We thank Sam Rickless, Eric Watkins, and an anonymous reviewer for their comments
on earlier drafts of this essay.