NICHOLAS STANG | Kant’s Modal Metaphysics | Oxford University Press 2016


By Jessica Leech

Kant’s Modal Metaphysics charts a fascinating course from Kant’s pre-Critical ideas about modality through to his more mature, Critical, view. We are not just offered an account of what Kant said about some narrow topic—modality—but rather a narrative according to which questions arising from Kant’s modal metaphysics play a crucial role in motivating and shaping the Critical philosophy. For example, in Chapter 6, it is proposed that questions of modal epistemology contribute to the Critical turn.

Given the importance of the pre-Critical ideas to this narrative, Stang devotes the first half of the book to them. In particular, much of Part I is taken up with reconstruction and discussion of the ideas and arguments that appear in Kant’s essay The Only Possible Argument in Support of a Demonstration of the Existence of God (henceforth Beweisgrund). This discussion plays an important role in the push on towards the Critical turn. The conclusion of the Beweisgrund argument, as Stang reads it, leaves significant questions unanswered, and raises important issues. According to Stang, it is these questions and issues that, in part, drive Kant’s thought onwards.

In this note, my aim is to examine Stang’s reconstruction of the modal argument of the Beweisgrund. The aim of the argument is to show that a simple, unique, absolutely necessary being exists (i.e. God). My discussion will be quite focused—on the reconstruction of one argument discussed in Part I of the book. However, given the important role played by this argument, its conclusion, and indeed the step of the argument on which I shall focus in Stang’s narrative of the development of Kant’s thought, my aim is not as narrow as it might seem.

First, I shall introduce some important definitions. Next, I sketch the modal argument as a whole. Then I move on to a closer examination of steps (ii) and (iv) of the argument. I cast doubt on the notion of an aggregate which is central to step (ii). Moving on, I suggest that existing interpretations of step (iv)—Stang’s included—overcomplicate matters, and offer what I take to be a simpler, more plausible, reading of this step. If I am correct, then some of what Stang takes to be the motivation from the Beweisgrund for the Critical turn is perhaps taken away, but by no means all. Significant questions of modal epistemology remain.

Absolute Necessity

In the Beweisgrund, Kant offers an argument for the existence of an absolutely necessary being. This is what Stang calls Kant’s modal argument.

Kant defines absolute necessity in a particular way, in terms of the grounding of possibility. Stang extracts three relevant definitions.

(Defn.) It is absolutely necessary that p if and only if ¬p cancels all possibility.

(Defn.) It is absolutely logically necessary that p if and only if ¬p cancels all logical possibility.

(Defn.) It is absolutely really necessary that p if and only if ¬p cancels all real possibility. (p. 124)

First, note that Stang leaves room for a distinction between necessity and absolute necessity. It might be necessary, but not absolutely necessary, that p. So these definitions concern only cases of absolute necessity. This also makes absolute necessity what is called a “hyperintensional notion” (p. 121ff.): in simplest terms, a hyperintensional notion can make distinctions amongst things that are necessarily equivalent. So, on Stang’s picture, amongst all the necessary truths we can draw a further line between those that are conditionally necessary (they are necessary because they follow necessarily from some other truths), and those that are absolutely necessary (they are necessary, not because they depend on other truths, but because they ground possibility in an important way). (One can similarly distinguish between necessary beings.)

Also note the distinction made between logical and real possibility and necessity. Stang discusses this distinction earlier in the book. In short, logical modality concerns lack of contradiction, real modality requires something more. Stang argues that the pre-Critical notion of real possibility requires that things be grounded in a particular (real) way. Thus, for the pre-Critical Kant (as well as for the Critical Kant), there are logical possibilities that are not also real possibilities, such as thinking matter (BDG, AA 2:85), and there are real necessities that are not logical necessities, e.g. that God exists (at least for the pre-Critical Kant).[1]

Setting the details of this distinction aside, most important for present purposes is the notion of absolute (real) necessity. Stang notes that, as things stand, the definition of absolute real necessity is not quite right. For it subsumes logical necessities under the real necessities. For example, “while the negation of the principle of contradiction cancels all real possibility, it does so by cancelling the logical element in real possibility” (p. 125). To define those necessities that are distinctively absolute and real, then, we need something slightly more specific:

(Defn.*) It is absolutely really necessary that p if and only if ¬p cancels the real element in all real possibility. (p. 125)

This is then cashed out as follows:

(1C) It is absolutely really necessary that p if and only if, were it the case that ¬p, there would be no real element in any real possibility.

(1E) For any x, x exists absolutely necessarily if and only if, were x not to exist, there would be no real element in any real possibility.

(1AP) For any x, x exists absolutely necessarily if and only if, were x not to exist, there would be no really possible atomic predicates. (p. 127)

There are, prima facie, two ways that something’s non-existence could thus relate to real possibilities. Its non-existence could cancel some real possibilities, or its non-existence could cancel all real possibilities. What should we make of this cancellation relation? Setting aside the details, the thought is that the thing’s existence should ground the real possibility, such that were that thing not to exist, that real possibility wouldn’t either. Hence, we are invited to think of these absolutely necessary beings as in some sense explaining the real possibilities. In the former case above, then, we have a ground of some real possibility (GSRP), and in the latter, a ground of all real possibility (GARP). The target of Kant’s argument is the existence of God as a GARP: as that unique absolutely really necessary being that grounds all real possibility.

Kant’s Modal Argument

According to Stang’s reconstruction, the Beweisgrund argument proceeds in 4 main steps:

(i) It is absolutely necessary that something exists.

(ii) There is an absolutely necessary being.

(iii) It is unique.

(iv) It is simple.

I have concerns about Stang’s reconstruction of (ii) and of (iv). In order to get to these steps, I shall have to also briefly sketch steps (i) and (iii).

Step i

First, grant Kant the definition of absolute necessity above (1AP), and grant that, if any (atomic) predicate is to be really possible, then it must have a ground for that possibility. Grant also that if some substance grounds the real possibility of an atomic predicate F, then the non-existence of that substance would cancel the real possibility of F. It follows that “if nothing were to exist, no atomic predicates would be really possible” (p. 129). But some atomic predicates are really possible. Hence, it is absolutely necessary that something exists. Note: this is the de dicto necessity that something or other exists, not the de re necessity that some particular thing is a necessary being.

Step ii

The second step effectively involves a move from GSRPs to a GARP, that is, from the existence of beings that are grounds of some real possibility to a being that is a ground of all real possibility. The idea is this. Suppose that it is absolutely necessary that something exists (step [i]). This doesn’t tell us that any particular being is absolutely necessary. For, suppose that x and y were both GSRPs. Then although, were x not to exist, the real possibilities grounded by x would be cancelled, there would still be some real possibility grounded by y. And mutatis mutandis for y and x. What would make some particular thing absolutely necessary would be if its non-existence would cancel all real possibility. That is, a GARP would be absolutely necessary. So step (ii) is intended to take us from GSRPs to a GARP.

How do we get there? We basically collect together all the GSRPs to make a big GARP:

If we take all of the beings that are grounds of some real possibility (GSRPs…) and treat them collectively as one being, then that being is a ground of all real possibility (GARP) and exists with absolute necessity. All really possible atomic predicates are grounded by some part of this collection and if this collection were to be completely annihilated (if all of its parts were annihilated) it would cease to exist. (p. 130)

However, Stang introduces a complication:

Some care is required here. We need this collection to be such that annihilating any proper part of it does not annihilate the collection itself; we need it to be such that the only way of annihilating it is to annihilate every one of its proper parts. (p. 130)

Why? Because if the GARP as a whole is a ground of all real possibility, then if it were to be annihilated, all real possibility would be cancelled. But its parts are not supposed to be GARPs in their own right—they are merely GSRPs. However, if by annihilating one part of the collection, one GSRP, one would thereby annihilate the whole collection, all real possibility would be cancelled. But this would mean that, after all, the part was a GARP, and not a GSRP.

What kind of a collection might be appropriate here then? Stang appeals to the notion of an aggregate, as introduced by Fine (1994).

The aggregate of the Xs exists if and only if at least one of the Xs exists. (p. 130)

What kind of a collection works like that? The examples drawn on by Fine involve temporal stages of things and events. For example, if one takes individuals to be four-dimensional beings with temporal parts, then an individual such as Stang can exist at time t1 although only a single temporal part of Stang exists then (his other temporal parts exist at different times). Or compare events: the birthday party exists at 10.01 pm although only one part of the party exists then (say, the blowing of the candles, and not the singing of ‘Happy Birthday’). Appealing to this notion of aggregate allows Stang to complete step (ii). He calls the aggregate of all GSRPs ‘Ω’, then appeals to the following principle:

(3) If F is a really possible atomic predicate and substance x grounds its real possibility, the non-existence of x cancels the real possibility of F. (p. 131)

Stang concludes:

By definition, Ω exists if and only if at least one GSRP exists, so the cancellation of Ω is equivalent to the annihilation of all GSRPs, which would cancel all really possible atomic predicates. So it follows that Ω exists with absolute necessity. Since Ω contains all GSRPs, it grounds all real possibility (it is a GARP). (p. 131)

As far as it stands, this does seem to yield the desired result of step (ii). However, one might entertain (at least) the following concerns.[2] It is an important part of Stang’s reconstruction of the argument here that one appeal to the right kind of aggregate. But is it plausible that Kant had access to such a notion, or that he had such a notion in mind? It is notable that this conception of an aggregate makes most sense, and is introduced by Fine, in terms of entities that have temporal parts. It is implausible that a notion understood in this way was part of Kant’s theoretical resources. Moreover, it also doesn’t appear to resemble the kind of aggregate relevant in the argument, which does not appeal to GSRPs as temporal parts of a GARP.

Stang anticipates these concerns in part. He specifies that his notion of aggregate is “a generalized non-temporal version of the Finean notion” (p. 130, fn. 17). However, it is unclear how we should understand this generalised notion. In the temporal case, we can make sense of the idea that temporally distant parts exist, although they don’t exist now, in the same way that spatially distant parts of things can co-exist, but not all in the same precise location. (E.g. the other end of my street exists, but not right here, just as the part of me that lived through 2016 exists, but not right now.) In the generalised case, we are asked to conceive of an aggregate such that only one of its parts exists at all, and the others do not. But then what is being aggregated? There seem to be (to exist) no other parts. One might introduce a modal analogue to the temporal and spatial examples: only one part of the aggregate actually exists, the others are modally distant, existing in other possible worlds. However, it is again implausible that Kant had this kind of sophisticated modally-extended entity in mind. Moreover, Stang has spent a good deal of Part I of the book arguing that Kant rejects the idea of possibilia and objects otherwise understood as non-existent.

Insofar as the crucial generalised notion of an aggregate remains unclear, then, this casts doubt on Stang’s reconstruction of step (ii). Until we can understand what these aggregates are, we can’t understand this step of the argument. The worry that Kant would not have had such a notion in mind also remains. Not only does this notion appear to make appeal to contemporary notions such as four- or five-dimensional objects, or possibilia, but it also appears to be in tension with other key views that Stang attributes to Kant.

Step iii

The aim of this step is to show that there can be only one GARP. Start by supposing that there are two distinct GARPs, A and B. Now, predicates are either fundamental or derivative. A fundamental predicate is instantiated by a GARP; the real possibility of a derivative predicate is otherwise grounded in a GARP (see pp. 103–4 and 131). Consider the fundamental predicates of A. The real possibility of these predicates is supposed to have no further ground than being instantiated by A. But as B is a GARP, the real possibility of these predicates is (also) grounded in B. Contradiction. Therefore, we should reject the starting premise, that there is more than one GARP.

Step iv

Having argued that our GARP Ω is unique, the final step is to argue that it is simple. Such an argument, apart from anything else, may go some way to alleviating any concerns about the aggregative nature of Ω suggested by step (ii). It should turn out that Ω has no parts after all, and so we are spared the worry of what its parts are like, and where they are. However, it is worth noting that this doesn’t entirely dispel my concerns. If the generalised notion of an aggregate plays a key role in the argument, then (a) we still need to understand the notion if we are to properly follow the argument, and so have any chance of being properly persuaded by it, and (b) we still need to assess whether the notion can be attributed plausibly to Kant’s conceptual repertoire.

Step (iv) turns on the following passage from Kant:

If one were to appeal to the definition of the necessary being and say that in each part the data of some inner possibility is given, but in all of the parts together all possibility is given, one would be imagining something wholly, albeit covertly, incoherent. For if one then thought that some inner possibilities could be cancelled, while others, given through other parts, remain, one would have to suppose that it is in itself possible for inner possibility to be negated or cancelled. But it is absolutely unthinkable and contradictory that something be nothing, and this means that cancelling any inner possibility eliminates all that is thinkable. It is apparent from this that the data for all that is thinkable must be given in the thing whose cancellation is the negation of all possibility; therefore, that which contains the ultimate ground of any inner possibility contains the ground of all possibility whatsoever, and this ground cannot be divided into distinct substances. (BDG, AA 2:84; translation taken from Stang, pp. 132–3)

The challenge is to try to make sense of this passage in a way that interprets Kant’s argument as ruling out ‘pluralism’: the view that all real possibility is grounded by a plurality of GSRPs, and that there is no genuine, simple GARP, but only a GARP in the sense of an aggregate of (more than one) GSRPs. Stang reviews a number of replies to this problem, which he calls the plurality objection, then offers his own.

I shall not reproduce the review of options here in detail, save to note that each involves some significant assumption or commitment beyond what obviously appears in the passage. The first option involves commitment to the S5 axiom (that all possibilities are necessarily possible) as well as a distinction between a counterfactual relation and a grounding relation. The second, Chignell’s solution, involves commitment to a real harmony principle, that “[i]f F and G are fundamental predicates and F and G are really compatible then the real compatibility of F and G must be grounded in their co-instantiation by a single substance” (p. 135). The third, Yong’s solution, appeals to the principle that relations between entities must be grounded by something. There are detailed defences of these ideas, and rebuttals from Stang. However, one might worry that none of these principles obviously appears in the argument passage above.[3] Hence, it would be preferable, all details aside, to find a way to understand the passage that did not depend upon principles which, whilst Kant may endorse them elsewhere, do not explicitly appear in the argument in which they are supposed to play a crucial role. After all, if the argument really rests on these principles, why wouldn’t Kant say so more clearly?

Stang’s proposed interpretation of the argument also draws on significant principles from elsewhere. He appeals to a weak principle of recombination:

It is really possible for there to be a world in which all the fundamental atomic predicates are instantiated (though not necessarily co-instantiated). (p. 138)

as well as a principle concerning the fundamentality of predicates:

If F is a fundamental predicate and x is the ground of the existence of something that instantiates F then x instantiates F. (p. 138)

and principles similar to that used by Yong:

If there exists a plurality of substances that really possibly interact then there exists a substance that is the ground of their existence. (p. 140)

The argument also requires appeal to a strong principle of sufficient reason (see pp. 140–2 for details of the argument). And finally, Stang argues that the notion of grounding employed in the argument must be understood in terms of God’s powers:

[D]erivative predicates are grounded, not in God’s actually causing them to be instantiated, but in God’s power to cause them to be instantiated. (p. 145)

Without going into the details of the argument, the same concern as above can be raised. Are those principles really used in Kant’s argument, even if they are available to him, that is, he endorses them elsewhere? Stang pre-empts such a concern when he writes that

[t]he point of this reconstruction, though, is not to show that Kant decisively demonstrated that there is a unique ground of all real possibilities. It is to show that he could have taken himself, plausibly, to have good reasons, from within his own metaphysical system, for asserting that possibility could not be ‘parcelled out’ among a plurality of grounds. (pp. 143–4)

If this is Stang’s aim—just to show that Kant has the resources to make such an argument—then one cannot criticise him for drawing on resources not explicitly in the passage. But we should then question how to understand the significance of this argument reconstruction within Stang’s wider narrative. If Kant didn’t in fact have this argument in mind, then it doesn’t seem right to use questions concerning this version of the argument to contribute to an explanation of the development of Kant’s thought.

It is from the end point of this argument—appeal to God’s power as a ground of really possible predicates—that Stang generates a tension for Kant. Elsewhere, Kant has argued that we cannot conceive of God grounding possibility through his powers (see p. 145). Stang writes:

This is the basic tension in Kant’s pre-Critical modal metaphysics: in order to understand his modal argument we need, implicitly, to think of God as grounding real possibility through his causal power, but we have independent reason to think this cannot be the case. (p. 146)

This leads on to a more general thought:

Beweisgrund, therefore, raises a question about our capacity to represent possibilities that it does not answer: how is it possible for us to represent the relation between the space of real possibility and its unique ground in such a way that would allow us to formulate and understand Kant’s own argument for the existence of such a ground? (p. 146–7)

This then leads us onwards to the Critical philosophy.

If all that we are to take from Stang’s reconstruction of Kant’s modal argument is that Kant had the conceptual resources to make this argument, but may not have done, then even if we can take there to be a conceptual progression from Stang’s version of the argument to the Critical philosophy, there would seem to be no reason to take this to be part of Kant’s own motivation for developing his thought, beyond more general considerations of modal epistemology. Indeed, it seems to me that a much simpler reading of the relevant passage has been overlooked.

Here is my proposed reading of the passage:

Suppose A is a ground of some, but not all, real possibilities, where all the GSRPs together collectively form one GARP. What distinguishes A from other GSRPs? They can come apart: if these parts are genuinely distinct, that means that some could exist without the others. That is, “some inner possibilities could be cancelled, while others, given through other parts, remain”. That is, A could cease to exist, and thereby cancel the possibilities that it grounds, while other GSRPs, say B, and their possibilities remain. But then this means that it is really possible for A not to exist, and hence for the possibilities that it grounds to fail to exist: “one would have to suppose that it is in itself possible for inner possibility to be negated or cancelled”. What could ground this possibility? (It is really possible, so it has a ground.) Not A. A couldn’t itself ground the possibility of its own non-existence. For then, in case that possibility became actual, there would be something grounding the possible non-existence of A that did not exist, i.e. which was nothing. But, “it is absolutely unthinkable and contradictory that something be nothing”. So, if A really could fail to exist, then this possibility must be grounded in something else, say B. But if the a-possibilities depend on the existence or non-existence of A, and if the possibilities for the existence or non-existence of A are grounded in B, then surely the a-possibilities are after all grounded in B, not A. So it is B, the “ultimate ground” for all of those possibilities, that is in fact the GARP. It cannot be divided into distinct substances, because the grounding done by any distinct substances collapses into grounding by one.

More formally:

1. Let Ω be a GARP, made up of distinct parts x1…xn, where each xi is a GSRP.

2. If x1…xn are distinct, then for each xi, it is possible that xi doesn’t exist and the remaining parts xj…xk exist.

3. ∴ For each xi, it is possible that xi doesn’t exist and remaining parts xj…xk exist [1,2]

4. If some xi is a ground of some real possibilities, its non-existence cancels some real possibilities (those it grounds).

5. ∴ It is possible that some real possibilities are cancelled, while some real possibilities are not cancelled. [3,4]

6. Every real possibility has some ground.

7. ∴ There is some ground for the possibility that xi doesn’t exist, and thereby some ground for the possibility that some real possibilities, but not all, are cancelled. [3,5,6]

8. For any x, x cannot ground the possibility of its own non-existence.

9. ∴ xi cannot ground the possibility that xi doesn’t exist. [8]

10. ∴ Some GSRP xj in GARP grounds the possibility that xi doesn’t exist. [1,6,9]

11. ∴ xj grounds the possibilities grounded in xi. [Transitivity of grounding]

12. ∴ xi is not a GSRP (xj ultimately grounds the real possibilities that we took to be grounded by xi.) [11]

13. ∴ Any part of Ω which is a distinct GSRP is not a GSRP. [1–12]

14. ∴ Ω has no distinct parts that are GSRPs. [1–13, reductio on 1]

This reconstruction of the argument does not obviously appeal to any significant principles that are not present in the passage. The main assumptions of the argument appear at (2), (4), (6), (8), and (11).

(2) is supported by the passage: “if one then thought that some inner possibilities could be cancelled, while others, given through other parts, remain”. (4) is part of Stang’s basic understanding of Kant’s thought about grounds of real possibility, which I am not challenging here. (6) is already assumed by Stang in step (i).

I take (8) to be supported by the passage:

[O]ne would have to suppose that it is in itself possible for inner possibility to be negated or cancelled. But it is absolutely unthinkable and contradictory that something be nothing.

As I read it, this is saying that one would have to suppose that it is possible for the inner possibility of something to be negated, not conditional on anything else, but in itself. But then this is tantamount to the thing itself grounding the possibility of it itself not being possible, and hence not existing. But then it would be possible for something to be nothing. And this Kant rejects in the passage.

(11) appeals to the transitivity of grounding, which seems to be a plausible principle of grounding. One might worry that this is yet another principle that doesn’t explicitly appear in the passage, and so I am falling foul of the same criticism I made of other options above.[4] In response to this, I would suggest that it appears to be such a simple and plausible principle that one might not deem it worthy of significant mention in the text, hence explaining why Kant makes no explicit appeal to it. This is in contrast to the more controversial, metaphysically heavyweight principles appealed to by other options. This is not to say that the transitivity of grounding is an uncontroversial issue—far from it.[5] But it is one of those principles that prima facie seems to be true, so that it is really quite interesting to find out that there may after all be problems with it. It doesn’t seem to me that the principles introduced in alternative interpretations of the argument are like this. In any case, I think it is also fairly plausible that Kant had something like the transitivity of grounding in mind when he mentions “the ultimate ground of any inner possibility” (emphasis added) in the passage, although I grant that this is not clear-cut.

In general, I offer a reading of the passage which appears to be an argument with steps, but which does not appeal to substantive principles that are not obviously mentioned in the passage itself. It seems to me to be a virtue if we can make sense of the argument without having to appeal to background principles, but rather just the proximal ideas of the Beweisgrund and the contents of the passage.

I suspect that Stang would object, amongst other things, that this version of the argument does not rule out a vicious regress of grounding: the possible non-existence of A, and hence the real possibilities grounded in A, are grounded in B, but then the real possibilities grounded in B may in turn, for similar reasons, be grounded in C, and so on and so forth (see p. 141). It is for this kind of reason that Stang introduces a principle of sufficient reason into play. However, it seems to me that there is a different response available for my version of the argument. The important work of the argument rests on the mutual ontological independence of the parts of the plurality (my premise 3), and the ground of the possibility of what it means to be independent parts—the real possibility of some existing without the others. Every time you try to introduce plurality, so the argument goes, it collapses into simplicity. This doesn’t generate a descending chain. It retains the single GARP argued for at step (iii), and collapses its purported parts into one.

The Critical Turn

In summary, then, my concern is that within Stang’s reconstruction of Kant’s modal argument we find the attribution to Kant of a problematic notion of aggregate, as well as the employment of principles that, whilst he does endorse them elsewhere, Kant does not obviously use in this context. As such, doubt can be cast on the extent to which we can understand any motivation for Kant’s philosophical revolution as arising from this version of the argument and its associated tensions or problems. That said, I find Stang’s more general claim about the role of modal epistemology to be an attractive one:

If not all logically consistent concepts are really possibly instantiated, by analysing our concepts and ascertaining their logical consistency we do not thereby come to know that they are really possibly instantiated. What then is the source of our knowledge of real possibility? This problem in modal epistemology poses a significant challenge to metaphysics, as Kant and his rationalist predecessors had practiced it. Ontology (metaphysica generalis) was conceived […] as the science of all possible beings qua possible beings. Consequently, a challenge to our modal epistemology, for both Kant and his predecessors, is also a challenge to the very idea that metaphysics is a science [Wissenschaft], a body of knowledge [Wissen], rather than merely rationally grounded conjectures or hypotheses. (p. 153)

Indeed, anyone who admits that there is more to modality than the logical faces these challenges. They should then also be interested in finding a solution, and thus should be interested in finding out about extant proposed solutions. In other words, they should all read Stang’s book.

Acknowledgements: Thank you to Damian Melamedoff, Andrew Stephenson, and Mark Textor for comments on previous versions of this piece.

Invited: 2 June 2016. Received: 5 January 2017.


[1] Thank you to Andrew Stephenson for help with these examples.

[2] There are other concerns one might have about this notion of aggregate, and this application of the notion, that I do not address here.

[3] See Stang (2016), Section 5.4, as well as Adams (2000), Chignell (2009, 2012, 2014), Watkins & Fisher (1998), and Yong (2014).

[4] Thank you to Andrew Stephenson for expressing precisely this worry.

[5] See, for example, Schaffer (2012).



Adams, R. (2000), ‘God, Possibility and Kant’, Faith and Philosophy 17(4): 425–40.

Chignell, A. (2009), ‘Kant, Modality, and the Most Real Being’, Archiv für Geschichte der Philosophie 91(2): 157–92.

——— (2012), ‘Kant, Possibility, and the Threat of Spinoza’, Mind 121(483): 635–75.

——— (2014), ‘Kant and the “Monstrous” Ground of Possibility: A Reply to Abaci and Yong’, Kantian Review 19(1): 53–69.

Fine, K. (1994), ‘Compounds and Aggregates’, Noûs 28(2): 137–58.

Schaffer, J. (2012), ‘Grounding, Transitivity, and Contrastivity’, in F. Correia & B. Schnieder (eds), Metaphysical Grounding: Understanding the Structure of Reality (Cambridge: Cambridge University Press).

Watkins, E. & M. Fisher (1998), ‘Kant on the Material Grounds of Possibility’, Review of Metaphysics 52(2): 369–96.

Yong, P. (2014), ‘God, Totality and Possibility in Kant’s Only Possible Argument‘, Kantian Review 19(1): 27–51.

© Jessica Leech, 2017.

Jessica Leech is a Lecturer at King’s College London. She has previously been a Lecturer at the University of Sheffield, and a Junior Research Fellow at King’s College, Cambridge. She did her doctorate jointly at the University of Sheffield and the University of Geneva (as part of the ‘Theory of Essence’ research project based at the Eidos Centre for Metaphysics at the University of Geneva). Leech’s research interests, contemporary and historical, centre around the topic of modality.