The idea that reality has a hierarchical structure is familiar and intuitive. For example, it’s natural to think that the psychological facts in some sense depend on and arise from the biological facts, the biological facts from the chemical facts, and the chemical from the physical. We find it natural to think of some states of affairs as more basic or fundamental than others, and to explain or account for one realm of facts in terms of others that are more fundamental. The grounding theorist—particularly, the kind of realist about ground I’ll be concerned with here—takes this hierarchical structure in a metaphysically serious way. That is, she takes reality to genuinely have such hierarchical structure, with the distinct levels of facts in this structure being related by a worldly relation of ground.¹ My focus in this paper is on the question of what exactly this metaphysical commitment entails. What is involved in being a realist about a worldly, metaphysical relation of ground?

I should note here that many grounding theorists take grounding to be an explanatory notion that is to be treated as an operator on sentences, rather than as a worldly relation between objects or facts.² And one can operate with the explanatory notion of ground without being committed to a worldly relation of ground or to any particular metaphysical structure. But my interest here is in the question of what’s involved in attributing grounding structure to reality, or what it takes for there to be genuine grounding in the world. My focus here will thus be on the worldly notion of ground. If we think there is a worldly metaphysical relation of ground that corresponds to the explanatory notion, or that we are attributing some distinctive metaphysical structure to reality when making grounding claims, we want to understand what that worldly structure must be like: So, what must the structure of reality be like to exhibit genuine grounding?

I’ll argue that there is a tension that is implicit in the notion of ground which makes it difficult to answer this question in a straightforward way. The tension is revealed via the attempt to make sense of the metaphysical status of the grounded—in particular, in the attempt to explain how it is that grounded facts can be distinct from and obtain in addition to their grounds on the one hand, and be “nothing over and above” their grounds at the same time. I’ll argue that straightforward accounts of the metaphysics of the grounded cannot satisfy both of these requirements. Either the higher-level facts are rendered too metaphysically separate from their grounds, or not separate enough—there appears to be no space for the metaphysical status of the grounded to be found.³

I’ll begin in Section 1 with an initial presentation of the puzzle, which reveals the tension I take to be implicit in the notion of ground in an intuitive way. I’ll then make the problem more precise in Section 2 and Section 3, and argue that the tension cannot be resolved as straightforwardly as it may appear. After rejecting some proposed solutions which I take to be unsatisfactory, I will present my preferred approach and explain how it can accommodate the metaphysical status of the grounded. The approach I present in Section 4 appeals to the notion of a perspective, and utilizes a meta-metaphysical framework that I have developed in application to other domains [Solomyak (2020); MANUSCRIPT]. I’ll argue that making sense of the metaphysics of ground requires that we embrace multiple perspectives on reality—corresponding to distinct ways or senses in which a fact can be said to obtain.

1 The Metaphysics of Ground: A Tension

The puzzle for the grounding realist can be brought out by attending to two aspects of the notion of ground, which I’ll call the two requirements of ground. I take each of these requirements to be essential to our intuitive understanding of what it is for one fact to be grounded in some further facts. But we’ll see that the two requirements pull in opposite directions, and reveal a tension in the notion of ground that causes trouble for the realist who wants to attribute genuine grounding structure to reality.

The two requirements of ground are what I’ll call Distinct Obtaining and Nothing Further:

Let’s start with Distinct Obtaining. First, why take [A] to be distinct from its grounds? To begin with, it’s important to note that on the above formulation (as on a very common conception of ground) [A] is a single fact, while its grounds is a plurality of facts.⁴ So we couldn’t generally take [A] to be identical to its complete grounds. And there are reasons to think that [A] must be distinct from each individual fact in the collection of its grounds as well: First, the grounded fact might have a structure or involve objects that aren’t involved in the individual facts that contribute to its grounds—a reason to think that the grounded fact is at least sometimes distinct from each of its (partial) grounds. But more generally, the facts in \(\Gamma\) are supposed to explain the obtaining of [A], or be an account of what makes it the case that A, and it’s not clear how a fact could explain or account for its own obtaining. If [A] appeared in its own grounds, the corresponding grounding explanation would strike us as circular. For these and other reasons, grounding is commonly taken to be irreflexive, and a grounded fact is taken to be distinct from any (and all) of the facts that ground it.⁵

Ground is also typically taken to be a factive notion.⁶ And on the factive notion of ground, [A]’s being grounded in \(\Gamma\) implies that [A] obtains. More generally, if some facts (which themselves obtain) make it the case that A, then it must be the case that A. It thus appears obvious and uncontroversial that grounded facts, in addition to their grounds, obtain. Distinct Obtaining, then, is implicit in a very common and intuitive understanding of what grounding involves.

Let’s now consider Nothing Further. Nothing Further is an expression of the tightness of the metaphysical connection that is supposed to hold between a grounded fact and its grounds. Unlike weaker relations such as supervenience, where one realm of facts can supervene on another while having a kind of metaphysically independent reality of its own, grounded facts can have no such metaphysically independent reality: the grounded facts are nothing over and above their grounds.^7,8 Of course, this “nothing over and above” may not be analyzable in terms that don’t ultimately appeal to the notion of ground.⁹ But the intuitive thought behind Nothing Further is that the grounded doesn’t carry any additional “metaphysical weight” once its grounds are in place. Given the grounds, nothing additional is required, metaphysically speaking, for the grounded facts to obtain. It is sometimes said that for a grounded fact to obtain just is for its grounds to obtain. For example, assuming that the existence of a table is grounded in the existence and arrangement of certain particles, we might say that for there to be a table just is for there to be particles arranged in this particular way.

But the “just is” in the above formulation is not meant to be the “just is” of identity—as we noted above, the intuitive notion of ground does not take the grounded to be identical to its grounds. And this is precisely where the worry for the grounding realist arises: What is the “just is” that figures in grounding claims that’s metaphysically tight enough to satisfy the requirement of Nothing Further, while still being short of identity? In other words, where is the space between “obtaining in addition to” and “obtaining over and above” for the metaphysical status of the grounded to be found? While the two requirements we’ve spelled out are both aspects of a single seemingly coherent notion of ground, they appear to pull in opposite directions: How can the grounded facts be distinct from and obtain in addition to their grounds on the one hand, and yet be “nothing over and above” their grounds at the same time?

One might think that simply accepting the notion of ground as a metaphysical primitive which exhibits the features described here is enough to allay the tension I’ve described: Perhaps it is precisely in being grounded that the metaphysical status of the higher-level facts meets both of the requirements above.¹⁰ The sense in which the higher-level facts are nothing over and above their grounds while still being distinct from them is in that the former are grounded in the latter. Ground is the primitive notion that exhibits the maximal-metaphysical-tightness-just-short-of-identity which we were after in the discussion above.

But this response avoids the real question at issue. Even for one who embraces ground as a metaphysical primitive, the question arises as to what structural features realist grounding claims implicitly attribute to reality. In what sense is a reality structured by a worldly relation of ground genuinely hierarchical? In what follows, I’ll argue that the apparent tension introduced above does in fact pose a real challenge for the grounding realist—one that taking the notion ground as a metaphysical primitive does not, in itself, suffice to meet. We’ll see that accommodating both requirements of ground in an account of reality’s structure is far from straightforward, and this will ultimately push us to reconceptualize our understanding of the metaphysics of ground—clarifying what it is to attribute primitive grounding structure to reality.

2 Two Pictures of Reality

Let’s begin by considering two different pictures of what the structure of reality might be like.¹¹ On the one hand, we have what I’ll call the Ground-Level Picture (GLP): On this picture, the ground-level, fundamental facts are ultimately all there is to reality—only the ground-level facts really obtain. Reality is thus ultimately “flat” rather than hierarchical on this picture—there is just one real level of facts.¹² On the other hand, we have what I’ll call the Hierarchical Picture (HP): On this picture, reality consists of both fundamental and non-fundamental facts. The non-fundamental facts really obtain, just as and in addition to their grounds. Reality thus has a hierarchical structure, with multiple “levels” of facts.

Of course, there is a question about how the uses of “really” and “ultimately” here should be understood. This is an issue we will return to shortly, and which will be central in the discussion that follows. For the time being, I want to appeal to an intuitive understanding of these locutions, as well as of the difference between the two pictures sketched above: We have a flat, ground-level-only reality on the one hand, and a hierarchically structured, multi-level reality on the other. We have an initial, intuitive understanding of what each of these pictures amounts to, and getting a more precise understanding of the commitments they involve will be our central aim in the discussion below.

What should the realist about ground say about these two pictures of reality? Must she be committed to one of these pictures over the other? At first approach, the Ground-Level Picture appears incomplete: The fundamental facts are not all the facts; the non-fundamental facts obtain as well. Given Distinct Obtaining—the first requirement of ground introduced above—the grounded facts are distinct from and obtain in addition to the facts that ground them. Further, the grounded facts are supposed to be those we “get for free,” or that “automatically arise,” once the ground-level facts are in place. All this seems to point against the Ground-Level Picture, and in favor of the Hierarchical Picture instead.

More generally, the Hierarchical Picture just seems to come along with the notion of ground, or more specifically, with realism about grounding as a worldly relation. And it’s a picture that is often presented as the grounding theorist’s background picture of the structure of reality. Here, for example, is Jonathan Schaffer’s description of the grounding theorist’s background theory:

[T]he neo-Aristotelian will begin from a hierarchical view of reality ordered by priority in nature. The primary entities form the sparse structure of being, while the grounding relations generate an abundant superstructure of posterior entities. (2009, 351)

So the realist about ground, who thinks of grounding as a worldly relation that structures reality, at least initially appears to be committed to the Hierarchical Picture.

But essential to this presentation of the two pictures is a non-trivial assumption about the difference between them. Implicit in the presentation of the two pictures as reflecting two distinct structures that reality might have, and in the conclusion that the realist about ground is committed to rejecting the GLP in favor of the HP, is what I’ll call the Assumption of Substantive Difference. This is the assumption that the GLP and the HP substantively differ with respect to the structure they attribute to reality. More specifically, it is the assumption that, according to the HP, reality has a metaphysical structure which is lacking according to the GLP.

Intuitively, this seems to be a very natural assumption. Whether reality is flat or hierarchically structured, and whether there are non-fundamental facts in addition to the fundamental facts, seem to be substantive questions about the structure of reality. Is the flat, Ground-Level Picture a complete picture of reality, or is it missing some of the facts that genuinely obtain? Are there distinct “levels” of facts? The GLP and the HP disagree over these questions and thus appear to reflect two distinct ways in which reality might be structured. And as we saw, the realist about ground appears to be committed to the Hierarchical Picture over the Ground-Level Picture, in maintaining that reality genuinely has the hierarchical structure that’s implicit in grounding claims.

But as we’ll see, this stance leads to a difficulty for the grounding theorist. In particular, we’ll see that the commitment to the HP, with the Assumption of Substantive Difference in the background, conflicts with Nothing Further—one of our original requirements of ground.

Recall that Nothing Further is the requirement that the obtaining of grounded facts be “nothing over and above” the obtaining of their grounds. Of course, we don’t have a precise way of cashing out how this “nothing over and above” should be understood, but we do have an intuitive sense of what it amounts to, and in what follows, I’ll argue that there is a real conflict between the intuitive “nothing over and above” that we require of the grounded and the assumption of Substantive Difference. It will thus turn out that favoring the Hierarchical Picture as the more accurate reflection of reality’s structure, as we think the grounding realist should, conflicts with an essential aspect of our intuitive notion of ground.

In the next section, I’ll consider some attempts to maintain the Assumption of Substantive Difference on behalf of the grounding realist and show why they are bound to be unsuccessful—arguing that Nothing Further and the Assumption of Substantive Difference are in fact in conflict, as I’ve claimed. I’ll then briefly consider some further responses to the puzzle, which I take to be unsatisfactory, and finally, in Section 4, present an alternative approach. There, I will argue that the Assumption of Substantive Difference is what must, after all, be given up. But realism about ground must thereby be significantly reconceptualized.

3 Maintaining the Hierarchical Picture

To see how the difficulty for the proponent of the Hierarchical Picture arises, we must ask ourselves what exactly is missing from the Ground-Level Picture that makes it incomplete, and which needs to be added in order to get the Hierarchical Picture. The Assumption of Substantive Difference says that there is something about reality that the Ground-Level Picture is missing out on, and the question is what that something could be. What could possibly be added to the GLP to get the HP, given that the obtaining of the higher-level facts is supposed to be nothing over and above the obtaining of the facts that ground them?

One might think that there is a trivial answer to this question: What needs to be added to the Ground-Level Picture is just all the higher-level facts. These facts obtain, and they are missing from the Ground-Level Picture. The problem with this response is that a proponent of the Ground-Level Picture will not deny that the higher-level facts obtain in a trivial sense. That is, someone who thinks that reality is ultimately exhausted by the fundamental facts does not deny that there is also an ordinary sense in which, (e.g.,) there are tables. The GL-theorist simply draws a distinction between the ordinary, trivial sense in which there are tables, and the further, somehow “metaphysically loaded” claim that there are really tables, or that this is a genuine fact of reality. For the GL-theorist, only the fundamental facts really obtain in this metaphysically loaded sense, but there is an ordinary, trivial sense in which the non-fundamental facts obtain as well.

This distinction relies on a metaphysically substantive notion of reality, or of what it is for a fact to really obtain. There are various ways in which such a notion might be cashed out, but to focus on one example, we might consider Kit Fine’s (2001) distinction between something’s being merely the case and something’s being the case in reality, where the latter has a metaphysical weightiness that the former does not. The GL-theorist might thus appeal to this distinction, and maintain that the non-fundamental facts obtain, but don’t obtain “in reality.” The GLP is a picture of what things are like in reality, and thus only includes the fundamental facts—the only facts that obtain in reality according to the GL-theorist.

The real difference between the Ground-Level and Hierarchical pictures, then, must be construed in terms of some further, non-trivial, metaphysical commitment—beyond the question of whether there are tables in an ordinary sense. The pictures disagree as to whether such non-fundamental facts really obtain in a metaphysically weighty sense —i.e., whether they obtain in reality. The Substantive Difference between the Ground-Level and Hierarchical Picture thus concerns the metaphysical status of the non-fundamental facts, not their mere obtaining in the most basic, trivial sense.

But on this understanding of the Hierarchical Picture, the obtaining of the higher-level facts appears to be something “over and above” the obtaining of the fundamental facts after all: If the metaphysical weightiness of the claim that the non-fundamental facts really obtain is something that could in principle be lacking even after the ground-level facts are in place, then the real obtaining of the higher-level facts consists in something more than the obtaining of their grounds. And this conflicts with Nothing Further: the requirement that the obtaining of the grounded facts be nothing over and above the obtaining of the facts that ground them.

It appears that once the ground-level facts are in place, and we agree that those really obtain, there can be nothing further at stake in the question of whether the higher-level facts really obtain as well. As soon as we admit that there’s a further metaphysical, or even meta-metaphysical, question to ask, an affirmative answer seems to grant the higher-level facts too much independent weight—the “real obtaining” of the non-fundamental facts in this metaphysically weighty sense is then something over and above the real obtaining of their grounds.

The grounding theorist might resist this line of thought by insisting that it doesn’t cost anything to posit the non-fundamental facts once the ground-level facts are in place—to say the non-fundamental facts are grounded in the fundamental facts is to say that we “get them for free” once we posit the obtaining of the fundamental. But given what I’ve argued here, it turns out that maintaining the commitment to Nothing Further requires more than this. It is not enough to say that we get the non-fundamental facts “for free”; rather, there must be no genuine difference, as far as reality is concerned, between a picture that includes them and a picture that doesn’t. There can be no metaphysically better answer to the question of whether they really obtain.

It appears, then, that the grounding theorist cannot maintain the Assumption of Substantive Difference by appealing to the higher-level facts, given her commitment to Nothing Further. Might the grounding theorist nevertheless maintain the Assumption of Substantive Difference in some other way? That is, might something other than the obtaining of the higher-level facts be what constitutes the Substantive Difference between the GLP and the HP? I’ll briefly consider two other possibilities.

One might attempt to appeal to facts about what grounds what in attempting to account for the difference between the two pictures. Perhaps it is not the higher-level facts themselves, but facts about how they are grounded that make for the Substantive Difference between the GLP and the HP. There are several ways in which this response could be developed, depending on how one understands the nature and status of these facts about ground, but as we’ll see, they all fall short for what is fundamentally the same reason.

On one view deRosset (2013), facts about ground are themselves grounded. On such a picture, appealing to facts about ground would be of no help to the grounding theorist: if the facts about ground are among the higher-level facts, they are simply a subset of those grounded facts we considered in the discussion above, which we argued could have no metaphysical weight over and above what is already in the GLP. The facts about ground—like any other higher-level facts—could not be the source for the substantive difference between the GLP and the HP given that their obtaining (like that of all higher-level facts) can be nothing over and above the obtaining of their grounds. Alternatively, one might consider a view on which the (or at least some) facts about ground are fundamental.¹³ But on such a view, the facts about ground would already be included in the GLP and thus could not be missing from it. They could not thus constitute the Substantive Difference between the GLP and the HP.

A third possibility for treating the facts about ground is suggested by Shamik Dasgupta (Dasgupta 2014), who argues that certain facts about how things are grounded are neither fundamental nor grounded, but rather, form a third category Dasgupta calls “not apt for grounding.” At first glance, such facts might be seen as a promising candidate for identifying the Substantive Difference between the GLP and the HP. But closer attention to this proposal reveals it to be unsatisfactory as well. We can define the GLP+ as the Ground-Level Picture together with the facts about ground, and then ask whether there is still something missing from the resulting picture. If the grounding theorist says there is something missing, we are back to our original question of what this could possibly be, given that the higher-level facts can be nothing over and above the facts that ground them. And if the grounding theorist says there is nothing missing, she appears to have rejected the Hierarchical Picture, which includes the higher-level facts as well as those at the ground-level.¹⁴ Irrespective of how we treat the facts about ground then, the puzzle for the grounding theorist remains: There is a tension between the grounding theorist’s commitment to the Ground-Level Picture over the Hierarchical Picture on the one hand, and the commitment to Nothing Further on the other.

More fundamentally, the problem with this series of proposals is as follows: The facts about ground—no matter where they are to be found in the grounding theorist’s metaphysical picture—concern the obtainment of the grounding relation between the higher-level facts and their grounds. Our central question has been how the metaphysical structure of this relation should be understood: What structure do we attribute to reality in making grounding claims? To say that grounding structure is accounted for by facts about that very structure is to get things upside down. These facts track the structure we are after rather than bring it about.

The grounding realist might attempt another avenue of response, and claim that the Hierarchical Picture is more accurate than the GLP holistically speaking, simply because it describes reality as layered, or hierarchical, rather than flat: the grounding relation is real, and genuinely structures reality. But this just sweeps the central question under the rug: What is it that makes the former a more accurate description of reality? If what makes it more accurate is the addition of the higher-level facts, we are left with the problem we encountered above —how could the addition of the higher-level facts make for a Substantive Difference, if their obtaining is nothing over and above the obtaining of their grounds? And if the Hierarchical Picture is more accurate for some other reason, the grounding theorist owes us an explanation of what that reason could be. The question we’ve been concerned with all along is the question of what is it for reality to have a hierarchical structure organized by the relation of ground. We’ve seen that there is a challenge to making sense of this position, and to simply restate the position does not help resolve it. How could the HP be a more accurate description of reality than the GLP given the requirement of Nothing Further? Absent some further theory or explanation from the grounding theorist, it is not clear how the Hierarchical Picture can be maintained, given the requirements of ground we started out with.

One might consider rejecting Nothing Further in favor of a weaker requirement on which it is allowed that grounded facts have metaphysical weight over and above that of their grounds. This would be to embrace a distinction between what we might call metaphysical cost and metaphysical weight, and to maintain that while it doesn’t cost anything, metaphysically speaking, for the higher-level facts to arise, their arising is in fact something over and above the obtaining of their grounds. The higher-level facts on such a picture would have a metaphysical weight of their own, in that their obtaining would be something over and above that of their grounds. Nevertheless, one would insist that this obtaining, or extra metaphysical weight, was of no metaphysical cost—the obtaining of the relevant grounds being all it takes for the additional layers of reality to arise.

This is perhaps closer to the Aristotelian picture of a hierarchically structured reality with multiple genuine “levels,” though I think it departs from a contemporary and very intuitive conception of ground, on which the grounding relation is supposed to be maximally “metaphysically tight.” But more importantly, I am suspicious of the conceptual move of distinguishing between metaphysical cost and metaphysical weight in this way. Unlike a buy-one-get-one-free deal at the supermarket—where one can get more “weight” than the “cost” one has paid—metaphysical weight cannot come free of metaphysical cost. Metaphysical cost and weight cannot, in principle, come apart: What it takes (metaphysically speaking) for a fact to obtain and what it is (metaphysically speaking) for a fact to obtain are one and the same. Any additional metaphysical weight that could in principle be lacking given the obtaining of the ground-level facts is thereby something it takes for the higher-level facts to obtain. The relevant notion of cost here is not causal, but metaphysical, and so there can be no metaphysical state or status that is lacking once the assumed metaphysical costs are in place.

The thought that one can maintain the Hierarchical Picture as a grounding theorist seems to implicitly rely on the mistaken assumption that the notions of metaphysical cost and metaphysical weight are separable: One says in the same breath that the grounded facts arise “for free” but also that they really do arise and thus constitute a genuinely hierarchical reality. As I’ve argued above, this is to grant the higher-level facts their own metaphysical weight, and thus requires that their weight can be separated from their cost. But closer attention to these notions once the distinction is made explicit reveals that there is no space for them to come apart. Metaphysical cost includes everything it takes for a fact to obtain, and this includes any metaphysical weight this obtaining may involve.

So I think Nothing Further is essential to the notion of ground, and should not be weakened or given up. But more importantly for our purposes, it’s clear that Nothing Further is essential to a very common and intuitive conception of ground, and I’m interested here in how we can make sense of realism about this particular notion.

To recap then, the puzzle for the grounding realist arises as follows: Distinct Obtaining, as well as general considerations surrounding the notion of ground, push against the Ground-Level Picture in favor of the Hierarchical Picture, implicitly endorsing the Assumption of Substantive Difference. But this stance is incompatible with Nothing Further, as Nothing Further implies that there can be nothing of metaphysical substance at stake in this move.

As I’ve argued, both Distinct Obtaining and Nothing Further are essential to an intuitive and very common conception of ground. Our only real option then, seems to be to reject the assumption of Substantive Difference—that the Ground-Level and Hierarchical pictures reflect two genuinely distinct structures reality might have. But this seems not only counter-intuitive in itself, but also counter to the intuitive notion of ground, which, as we’ve seen, naturally comes along with a Hierarchical Picture of the structure of reality.

In what follows, I’ll argue that the Assumption of Substantive Difference is nevertheless what the grounding theorist should give up. I’ll present a resolution to the puzzle on which hierarchical structure is understood to be compatible with—in fact, even dependent on—the rejection of Substantive Difference.

4 The Perspectives Approach

At this point, one might find it natural to react to the line of argument I’ve put forward as follows: Perhaps we’ve simply misinterpreted both the Ground-Level and Hierarchical Pictures to begin with in thinking that they could be pitted against each other; the Assumption of Substantive Difference should have been rejected from the start. Just as we concluded that the Ground-Level Picture shouldn’t be taken to deny the higher-level facts in a trivial sense, the Hierarchical Picture shouldn’t be interpreted as granting them any metaphysical weight of their own. The vertical arrangement of the facts in the hierarchy, as well as the arrows going from one level to the next, are precisely meant to convey that the relation between the lower and higher-level facts is one of grounding, and that the higher-level facts are nothing over and above their grounds.

I think this line of thought is correct, but its implications must be fully appreciated. To embrace this line of thought, one must admit that each picture taken on its own is deeply misleading. The Ground-Level and Hierarchical Picture turn out to be inter-dependent: In order to interpret each picture correctly, we must have the other picture in the back of our minds. To interpret the Ground-Level Picture correctly, we must see the non-fundamental as implicitly arising from the fundamental, as depicted in the Hierarchical Picture; and to interpret the Hierarchical Picture correctly, we must see the non-fundamental facts as not really anything over and above what’s already present in the Ground-Level Picture. Both pictures are thus essential to our conception of a reality that’s structured by the relation of ground.

This is not to say that the grounding theorist cannot speak of a hierarchically structured reality, or that reality cannot truly be hierarchically structured. But it turns out that what it is for reality to be hierarchically structured (on this understanding) is not at all what we would have thought. To say that reality is hierarchically structured is not to say that the Hierarchical Picture (as defined above) is the picture that best reflects reality’s structure. A truly hierarchical structure is, paradoxically, one that is best reflected by both the Hierarchical and the Ground-Level Pictures taken hand-in-hand —the hierarchical aspect of the multi-level structure is only guaranteed by the fact that the flat, ground level of the hierarchy is in some real sense all there is to the “hierarchy” at all.

In what follows, I want to suggest a way of conceptualizing this seemingly paradoxical state of affairs—one that I think can help clarify the notion of ground and further illuminate what the nature of a hierarchically structured reality must be like. The approach relies on the notion of a perspective, and distinguishes between two perspectives on reality implicit in the notion of ground. On the one hand, there is what I’ll call the ground-level perspective, from which reality is exhausted by the fundamental, ground-level facts. On the other hand, there is what I’ll call the hierarchical perspective, from which reality extends beyond the fundamental to encompass the higher-level, non-fundamental facts as well. As I’ve already suggested, these two perspectives are each essential to the notion of ground as well as to a reality that is genuinely structured by a worldly grounding relation. In what follows, I’ll present the background framework of perspectives in more detail, and then return to explain how this approach can provide a satisfying resolution to our puzzle.

4.1 The Perspectives Framework

The notion of a perspective that I appeal to here will remain undefined. But an intuitive gloss and a few examples will help bring out the particular notion of a perspective that I have in mind. To get an initial sense for the relevant notion, we can consider the familiar shift from seeing reality from a first-person, or subjective perspective, to seeing or conceptualizing reality in an impersonal, or objective way. Imagine an extreme solipsist who is not aware that there is a reality beyond her own subjective experience at all. Such a solipsist implicitly identifies her own experience with the whole of reality; for this solipsist, there is no distinction between something’s being the case in her experience and something’s being the case, full stop. Implicitly first-personal claims such as “it’s painful” or “it’s pleasant” will have absolute truth-values for this solipsist; from her perspective, how things are in reality and how things are in her experience are one and the same.

This strong identification of one’s own experience with the whole of reality is what I call the first-personal perspective, and can be contrasted with the broader impersonal perspective, from which one recognizes that reality extends beyond one’s own experience to include other subjects and/or objective states of affairs. From the impersonal perspective, a distinction is drawn between something’s being the case in one’s own experience and something’s being the case, full stop. First-personal claims such as “it’s painful” will (from this perspective) be incomplete without reference to a subject—things can be painful for one subject but not for another, and more broadly, how things are in one’s experience and how things are in reality can come apart.

Crucially, the shift from the first-personal to the impersonal perspective involves a change in one’s conception of reality, and more specifically, one’s conception of what we might call the “shape” of reality: from the first-personal perspective, reality is implicitly taken to be first-personal, while the impersonal perspective takes reality to be broader and “impersonal” in shape.¹⁵

We can similarly identify distinct perspectives on reality we might adopt in thinking about the metaphysics of time. On the one hand, there is the perspective of the present, or the present-tensed perspective. From this perspective, one identifies the present with the whole of reality. Crucially, this is not just an ontological stance. Rather, it is a more general conception of reality, which also includes a conception of what it is for something to be the case in reality, or of what it is for a fact to obtain. From the present-tensed perspective as I understand it, for something to be the case and for something to be the case now are one and the same. There is no metaphysical distinction to be drawn between something’s being the case in reality and something’s being the case in the present.

On the other hand, we can shift to the broader atemporal perspective, from which reality is seen as extending beyond the present to encompass other times and/or atemporal states of affairs. Again, this is not just a matter of ontology. From the atemporal perspective, one’s conception of reality allows for something to be the case in reality, but not in the present—e.g., something can be the case at another time, or just independently of how things are in the present. And here, just as in the first-person case, we can see ourselves as shifting from one of these perspectives to the other, and as shifting from a narrow to a broader conception of reality when we make that move.

In each case, we can ask whether one perspective or the other is more fundamental or metaphysically privileged as a perspective on reality. That is, we can ask whether it is the narrow or broad conception of reality in each case that is getting the shape and structure of reality “right.” Is reality ultimately first-personal or impersonal? Present-tensed or atemporal? The question of which, if either, perspective in each case is fundamental is a way of getting at this question about reality’s structure.

More generally, then, a perspective is a way of conceptualizing all of reality, and comes along with a corresponding conception of what it is for something to be the case in reality, or of what it is for a fact to obtain. That is, in adopting a perspective, one identifies reality in a certain way, which allows one to answer certain questions about what it is for a fact to obtain, as well as about which facts obtain, just by virtue of one’s conception of the metaphysical “shape” of reality. For example, in taking reality to be present-tensed, as one does from the present-tensed perspective, one thereby rules out non-present facts as well as objects from one’s ontology—one’s identification of reality with the present entails that what it is for a fact to obtain and what it is for a fact to obtain now are one and the same, and thus that obtaining at another time, or obtaining independently of time, are not ways of obtaining in reality. A perspective is thus associated with a certain answer to the question of what it is to be real, or of what it is to obtain in reality, from which ontological and other metaphysical commitments follow downstream.

In taking a perspective to be fundamental, one takes that perspective to be metaphysically privileged in its reflection of reality’s “shape.” A fundamental perspective is one that identifies the shape of reality correctly—such that what it is to be the case from that perspective and what it is to be the case, full stop, are one and the same. In other words, the “way of obtaining,” which this perspective identifies with “obtaining in reality,” correctly reflects what it is for a fact to obtain.

With this brief introduction to the perspectives framework in hand, we can return to the case of ground, and see how we might reconceptualize the surrounding metaphysical issues. As I’ll argue, adopting the language of perspectives allows us to resolve the puzzle for the grounding realist in an intuitively satisfying way.

4.2 Perspectives and Ground

Turning back now to the grounding case, we can see the Ground-Level and Hierarchical Pictures as corresponding to two distinct perspectives in the sense introduced above. The ground-level perspective is the perspective from which the fundamental level is identified with the whole of reality. From this perspective, for something to be the case and for something to be the case fundamentally are one and the same; no metaphysical distinction is drawn between fundamental reality and reality. On the other hand, we can shift to the broader hierarchical perspective, from which reality is seen as extending beyond the fundamental to encompass the non-fundamental as well. From this perspective, for something to be the case and for something to be the case fundamentally are not one and the same; something can be the case, but not fundamentally.

As in the first-person and temporal cases, we can see ourselves as naturally shifting from one of these perspectives to the other: We can conceive of reality as exhausted by the fundamental, and then broaden our conception of reality to include the non-fundamental as well. Importantly, this is not just a matter of “adding facts” into our picture of reality; rather, it involves shifting our conception of reality from a narrow sense to a broader one. We can grasp a sense of “reality” on which reality just is fundamental reality—to be real and to be fundamental are (on this conception) one and the same. On the other hand, we can also grasp a broader sense of “reality,” on which reality and fundamentality can come apart. Each of these perspectives, or conceptions of reality, thus comes along with a corresponding conception of what it is for a fact to really obtain: From the ground-level perspective, for a fact to really obtain is for it to obtain fundamentally, while from the hierarchical perspective, a fact can really, but not fundamentally, obtain.

We can now understand the move from the Ground-Level to the Hierarchical Picture in a new way. Rather than holding a fixed conception of reality and positing it to include additional facts, we are shifting our conception of reality. The GLP understands what it is to really obtain in one way, while the HP understands it in another. This shift in the way reality is identified automatically gives rise to “more facts” in the Hierarchical Picture than are present in the Ground-Level Picture —but this is because the criteria for being a “real fact” have been changed, not because these facts have been granted a heftier metaphysical status. The two pictures (now understood to be perspectives) thus correspond to two different ways, or senses, in which facts can be said to obtain.

The disagreement between the Ground-Level and Hierarchical Pictures, when seen as pitted against each other, can be seen as a disagreement about which of these two perspectives is fundamental or metaphysically privileged in its identification of the “shape” of reality: The GL-theorist holds that a conception of reality on which reality is identified with the ground-level is what best captures reality’s structure, while the proponent of the Hierarchical Picture takes the broader conception of reality to be fundamental—i.e., metaphysically privileged in its reflection of what it is to be real.¹⁶

But as the challenge I’ve raised for the grounding theorist illustrates, both perspectives are essential to accommodating the requirements of ground: Distinct Obtaining, as well as the general thought that reality has a hierarchical structure, requires that we adopt the hierarchical perspective, and recognize a sense in which the non-fundamental facts really obtain. Nothing Further, on the other hand, requires that we adopt the ground-level perspective, and recognize that anything “beyond” the ground-level is really nothing at all—i.e., nothing over and above what is already there at the ground-level.

Recognizing this way in which we implicitly adopt and shift perspectives is the key to resolving the puzzle for the realist about ground. For the grounding realist, the GLP and the HP are not to be seen as distinct ways in which reality might be structured; rather, they are to be seen as reflecting the two distinct perspectives that are both implicit in and essential to the notion of ground. Distinct Obtaining and Nothing Further are satisfied via these two distinct, but inter-dependent perspectives: Distinct Obtaining is satisfied by the fact that the hierarchical perspective is a “genuine perspective on reality”—i.e., that there is a real sense in which both fundamental and higher-level facts really obtain. Nothing Further, on the other hand, is satisfied by the fact that the ground-level perspective is a “genuine perspective on reality” as well—i.e., by the fact that there is a real sense in which reality is exhausted by the fundamental.

It is important to distinguish between a number of different claims that embracing this approach could involve. First, there is the weaker, conceptual claim that both perspectives are essential to our grasp of the notion of ground. That is, we might say that our grasp of the notion of ground relies on our ability to adopt both of these perspectives and shift back and forth between them, grasping two distinct senses of reality as we make that move. As I’ve said, I think we do in fact implicitly shift perspectives in this way, and making this explicit can help make sense of our conflicting intuitions in this area: To see the higher-level facts as genuinely grounded in the fundamental, we must think of them as really obtaining in one sense, and as nothing “beyond” what really obtains in another.

So I think the conceptual claim goes some way towards clarifying the issues surrounding the notion of ground. But our central challenge has been the metaphysical question of what it would take for reality to genuinely have the kind of hierarchical structure one commits to by being a realist about ground. And this brings us to the stronger, metaphysical claim which the grounding theorist might be pushed to embrace: namely, that the two perspectives in this case are not only essential to the concept of ground, but also to the metaphysical structure of a reality that exhibits genuine grounding. More specifically, this would entail a commitment to a kind of pluralism about reality, on which there are genuinely two distinct ways in which a fact can be said to obtain. For such a pluralist, there would be no univocal answer to the question of whether the higher-level facts really obtain: only the fundamental facts really obtain in one sense, while both fundamental and non-fundamental facts really obtain in another.¹⁷

Importantly, this pluralist stance must be distinguished from a kind of semantic pluralism on which the GLP and the HP simply reflect two different ways of speaking about the same reality, with no further fact of the matter as to which of the associated perspectives is fundamental. Embracing something like quantifier variance, or different—equally good—senses of the word “fact” or “obtains” would be other ways of rejecting the Assumption of Substantive Difference, but to embrace such a stance would be to give up on the robust metaphysical realism about ground that is of essential interest to us here:¹⁸ A stance on which there is no metaphysically privileged way of answering the question of how reality is truly structured is thereby also an anti-realism about ground as a relation that genuinely structures reality.¹⁹

The kind of pluralism I present here is thus more radical than it appears, and faces a number of conceptual and metaphysical difficulties, but I think it is in some ways best suited to reflect the commitments of the realist about ground. In what follows, I’ll sketch the proposal in a bit more detail, and explain how it can accommodate genuinely hierarchical structure.

4.3 Maintaining Hierarchical Structure

I’ve argued that there are two perspectives that are implicit in our thinking about the metaphysics of ground: the ground-level perspective, from which reality is identified with the fundamental, and the hierarchical perspective, from which reality is seen as extending beyond the fundamental to encompass the non-fundamental as well. A pluralist about reality takes each of these perspectives to correspond to a real way of being the case, or sense in which facts can obtain. Only the fundamental facts really obtain in one sense, while both fundamental and non-fundamental facts really obtain in another. For the pluralist, there is thus no univocal answer to the question of whether the higher-level facts really obtain in addition to their grounds: from the ground-level perspective on reality, they do not, while from the hierarchical perspective, they do.

It is important to note that on this approach, no distinction is drawn between something’s being merely the case and its being the case in reality, as it is on Fine’s (2001) view. The pluralist has no need for this additional distinction, given that she accepts multiple senses in which facts can (really) obtain in the first place. The “really” in the pluralist’s claim that the non-fundamental facts “really obtain” in one sense but not in another is thus metaphysically redundant—it only serves to make clear that, from the relevant perspective, what we are concerned with is as “hefty” a metaphysical status as there is. For the pluralist, there are simply two such metaphysically privileged statuses—i.e., two senses in which a fact can (really) obtain.²⁰

It is also important to make clear that the pluralist needn’t maintain that the two perspectives at issue here are metaphysically on a par, or that they are both maximally fundamental. It may be that one of the two perspectives is more fundamental than the other, but that both are still metaphysically privileged in that they each truly reflect something about reality’s structure. There are several options one might pursue here.

One might find it natural to think that the ground-level perspective is more fundamental than the hierarchical perspective, or even that it alone is maximally fundamental, while the hierarchical perspective is not. The latter stance has the significant benefit of bringing together the two distinct senses of fundamentality I’ve been employing here: the perspective that is fundamental qua perspective is taken to be the perspective that “sees” only the fundamental level, i.e., from which only fundamental facts really obtain. The hierarchical perspective would then be seen as non-fundamental, though essential to making sense of the metaphysics of ground. Though such a stance would be natural for the grounding theorist to adopt, it carries the odd and somewhat counter-intuitive consequence that, in the fundamental sense, hierarchical structure wouldn’t be real after all. Realism about ground is in some sense rendered impossible on this view—reality cannot really be hierarchically structured or organized via the relation of ground, in that grounding structure is only “visible” from a non-fundamental perspective on reality.

Importantly, this is not just the trivial claim that grounding structure cannot be found at the ground-level. One might find it natural to think that despite grounding structure’s not being fundamental in this sense, it is still real in a fundamental sense of the term. To maintain that grounding genuinely structures reality (i.e., that it’s real in a fundamental sense, even if not itself fundamental), one must take both the hierarchical and ground-level perspectives to be fundamental as perspectives—i.e., to correspond to genuine, “joint-carving” conceptions of reality.

But as I said above, this still leaves several options open, in that one might take the two perspectives to be equally but not maximally fundamental, or take one perspective to be more fundamental than the other—depending, of course, on whether one allows for a comparative notion of perspective-fundamentality. Developing a more thorough conception of perspective-fundamentality, as well as of the various routes one might take here, are beyond the scope of this paper. My aim here is just to provide a rough sketch of the picture, which I think most faithfully reflects the grounding theorist’s commitments, as well as of the resolution such a picture could offer for the realist’s puzzle. The intuitive appeal of the solution it offers can motivate the further development of variants of the pluralist view.

Nevertheless, it’s important to flag that the questions surrounding perspective-fundamentality and the precise formulation of the pluralist’s view raise some serious conceptual and metaphysical challenges. For one thing, it is not clear how the view can be formulated without reliance on a third sense of “reality”—in saying that there are two distinct, joint-carving conceptions of reality, we seem to employ a third sense in which this can be. There are various routes one might take in response to this worry. One is to claim that the third sense of “reality” is simply non-fundamental, and that the two that the pluralist has identified are simply the two that best “carve at the joints.” Another is to say that the third sense is more or even maximally fundamental, but not one we ordinarily employ. I explore each of these options elsewhere [in my Solomyak (2013); MANUSCRIPT], and wish to remain neutral on this issue here. For our purposes here, we can take the pluralist’s claim to simply be that there are two metaphysically privileged conceptions of reality, corresponding to two distinct ways in which facts can be said to obtain.

With this claim in hand, the grounding theorist can make sense of hierarchical structure in a way she was previously unable to do. What it is for reality to have hierarchical structure, on the pluralist’s understanding, is the following: In one real sense, there are both fundamental and higher-level facts, and in another real sense, reality is exhausted by the fundamental. The fundamental facts are thus real no matter which perspective one adopts, while the higher-level facts are real in one sense and unreal in another. This secures both the special status of the fundamental facts and the distinctive status of the grounded. The special status of the ground-level facts is secured by the fact that in addition to obtaining as the higher-level facts do, there is a real sense in which they exhaust reality; while the status of the grounded facts is secured by the fact that they really obtain in one sense but not in another—Distinct Obtaining secured by the fact that there is a real sense in which they obtain, and Nothing Further secured by the fact that there is a real sense in which they do not. For the pluralist, obtaining in one real sense but not in another is (part of) what it is for something to be a grounded, higher-level fact.²¹

One might worry that there is a sense in which the reality of grounding structure is not secured after all, in that one of the essential requirements of ground is not satisfied from each of the two fundamental perspectives on reality. The two perspectives are incompatible, and each is missing out on an essential element of the metaphysics of ground. How is it that the pluralist gets to satisfy both requirements by embracing the two perspectives rather than, ultimately, neither (or at least, always not-both)?²² To provide a fully satisfying response to this worry would require answering the questions posed above about the precise formulation of the pluralist view and the complications this raises. In particular, what is the more general sense of “reality” we are to use in stating that there are two fundamental perspectives? It is in this sense that the pluralist can say that grounding structure is real—each of the two requirements of ground is satisfied from one of the two fundamental perspectives. The objector might press on: Can we not equally say that each of the two requirements is, from some fundamental perspective, unsatisfied? But the intuitions behind the requirements of ground do not demand that each requirement be met in every sense, or even in every fundamental sense. What is essential is that there is some real sense in which the grounded obtains in addition to its grounds, and some real sense in which it is nothing over and above these grounds. It is thus enough that each requirement is satisfied from some fundamental perspective. This allows for hierarchical structure to be “real” in the more general, and yet to be fully explicated, sense of the term.²³

So this, then, is how the metaphysical status of the genuinely grounded is to be found. To be grounded is not to belong to a mysterious in-between state, wedged between “obtaining in addition to” and “being nothing over and above.” Rather, the grounded is what in one sense obtains in addition to, and in another sense, is nothing over and above. What makes the structure of reality genuinely hierarchical is that the “higher-levels” are, in one sense, there and, in another sense, literally nothing over and above the fundamental.²⁴

While what I have presented here is far from a developed metaphysical picture, it can be taken as a rough sketch of the direction in which I think the puzzle for the grounding realist pushes. The intuitive appeal of the pluralist’s solution can bring out the ways in which we implicitly adopt and shift perspectives in thinking about the metaphysics of ground, and bring to light what simpler solutions to the realist’s puzzle are thereby missing: A univocal answer to the question of whether the grounded facts really obtain cannot fully accommodate the competing requirements of ground. Recognizing that there are two genuine perspectives on reality here is the key to making sense of genuinely hierarchical structure, and developing a more thorough framework within which this claim can be understood is my aim elsewhere.

5 Conclusion

We’ve seen that the realist about ground faces a difficulty in accommodating genuinely hierarchical structure. The notion of ground requires that the grounded facts obtain, distinct from and in addition to their grounds, and that they be “nothing over and above” their grounds at the same time. These requirements push the realist about ground in seemingly competing directions: The realist is pushed to accept the Hierarchical Picture over the Ground-Level Picture on the one hand, and to reject that there is any Substantive Difference between these two pictures on the other. The higher-level, grounded facts are thus mysteriously elusive in their metaphysical status; we find them either too “metaphysically weighty” to be grounded, or too “metaphysically light” to support the Hierarchical Picture which seems essential for the grounding realist to maintain.

I’ve argued that the best route for resolving this puzzle involves rejecting what I’ve called the Assumption of Substantive Difference; The Ground-Level and Hierarchical Picture are not to be seen as reflecting distinct ways in which reality might be structured after all. But appreciating the implications of this stance and making it intuitive require some further reconceptualization of the issues. To make sense of the way in which the two pictures do not substantively differ, we can adopt the framework of perspectives: From one perspective, reality is exhausted by the fundamental, and from another perspective, reality encompasses the grounded, non-fundamental facts as well. The deep pluralism of this stance can provide a unique approach to understanding the metaphysical status of the grounded: the grounded is in one sense real, and in another sense literally nothing over and above the fundamental.