Why Critical Social Ontologists Shouldn’t Be Univocalists




How to Cite: Harris, J. (2021) “Why Critical Social Ontologists Shouldn’t Be Univocalists”, Ergo an Open Access Journal of Philosophy. 8(0). doi:

Critical social ontology is a burgeoning enterprise in the philosophy of social science that amounts to something like a synthesis of two seemingly disparate traditions of inquiry: namely, “critical theory,” broadly construed, and “ontology” or “metaphysics.” Like critical theorists, critical social ontologists understand their theoretical work to be formally conditioned by the fundamental project of human emancipation and social transformation.1 And, like metaphysicians, critical social ontologists concern themselves with questions about the being of the things they investigate.2 Following Michael J. Thompson, we might stipulatively define critical social ontology as a theoretical enterprise founded upon the conviction that “normative claims [e.g., claims about social injustice, oppression, etc.] are critical only to the extent that they carry descriptive claims about the essential structures of sociality within them” (2017: 18). In other words, the basic idea is that we need to attend to the distinctive being, essences and powers of social phenomena if we are to account adequately for their effects, especially those bearing on matters of social justice. Simply put, as Emmanuel Renault has noted, “different ontological assumptions can lead to distinct conceptions of social transformation” (2016: 29).

This being the case, I take it as obvious that, among other things, critical social ontologists should affirm what Katherine Ritchie (2015) calls “group realism,” that is, the thesis that social groups exist, and are ontologically irreducible to their individuals and their properties. Ritchie captures the basic line of reasoning I am after: “[b]eing part of [e.g.] a racial group can substantially affect one’s experiences, what one can do, and how one is treated. … Anything efficacious exists, so the argument goes, groups exist” (2015: 311). Indeed, I take it that philosophers such as Sally Haslanger, Charlotte Witt, Elizabeth Barnes, Tracy Isaacs and Átsa (among others) affirm Group Realism on relevantly similar grounds, that is, in a way that is consonant with what I (following Thompson, Groff and Renault) am calling critical social ontology.3

So, while a strict definition is beyond the scope of this paper, it is at least clear that critical social ontology is marked in these three ways: namely, its (1) constitutive “concern for social transformation” (Horkheimer 2014: 241); (2) commitment to ontological analysis of social entities; and (3) group realism. Therefore, in the interest of lending further theoretical support for these core commitments, I argue in this paper that critical social ontologists would do well to devote some attention to a heretofore (to my knowledge) unasked question in this literature: namely, the question of whether the predicate ‘exists’ is univocal as said of (a) social groups; and (b) other, less controversially existing things. In other words, does the predicate ‘exists’ have exactly one literal meaning, or are there literally various, though related meanings that have various “modes” of existence as their referents?4

Drawing from some traditional sources and recent work in metaphysics, I defend the latter view and explore its implications for group realism, using a recent dispute between Michael Schmitz and Raimo Tuomela as a case study.5 In short, my thesis is that an uncritical presumption of univocalism about existence may be “holding back” otherwise robustly realist accounts of social groups—and that it does not have to be this way. A non-univocalist account of existence provides a “clearing,”6 of sorts, in which social groups can reside, plausibly and without injury to reasonable parsimony constraints.

Critical social ontologists think that social things do things in the world. Therefore, at least to the extent that critical social ontologists are inclined towards admitting the literal existence of social groups, I claim that they should not be univocalists.

1. What Is at Stake? Univocalism and the Aristotelian Alternative

It is a core conviction of the Aristotelian metaphysical tradition that “being is said in many ways” (Aristotle 1967: 1003a). Call this the Aristotelian thesis. Of course, like any other core philosophical thesis, exactly what it entails is a subject of considerable controversy—even and especially within the gates of this long tradition. For my purposes, however, it is enough to identify some of this thesis’s uncontroversial implications, since even the uncontroversial ones stand in stark contrast to prevalent metaphysical intuitions formed at least in part by the combined legacy of philosophers such as Frege, Russell and Quine. Here are two of those implications, semantic and ontological, respectively:

Semantic Implication: ‘Exists’ has different meanings when predicated of subjects belonging to different ontological categories7 (e.g., if I say “I exist” and “the number 2 exists,” I don’t mean ‘exists’ in exactly the same sense).

Ontological Implication: Existence comes in a multiplicity of “modes” (e.g., my existence is distinct, though related to the existence of the number 2).

For current purposes, at least, I have these implications in mind when I refer to the Aristotelian position. Conversely, by univocalism, I mean the respective contradictory opposites of these implications. In other words, if univocalism is true, then the Aristotelian thesis must be false, and vice versa.8

Now my main concern in this paper is to consider the difference this Aristotelian alternative might make with respect to the real existence of social groups. I cannot adequately defend a general account of what social groups are in this paper, but suffice it to say that I have in mind something like the account recently developed by Ritchie: namely, that social groups are “structured wholes” in which individual persons occupy “nodes,” either intentionally or unintentionally.9 Paradigmatic examples of such groups include informal and formal committees, teams and firms, as well as class and racial groups.

So, while I am not concerned primarily with general metaphysics per se here, I will nevertheless address some of the primary motivations for these competing positions—if only to avoid appearances of special pleading on behalf of the Stagirite’s views.

1.1. Motivations for Univocalism

Although what I am calling univocalism about existence precedes W. V. O. Quine,10 he is nevertheless a lucid and influential representative of the doctrine. In his essay, “On What There Is,” he famously says that “To be assumed as an entity is, purely and simply, to be reckoned as the value of a variable” (Quine 1961: 13). In other words, to say “x exists” is simply to say “Some x is F,” and nothing more (Quine 1961: 12).

There are a number of attractive features of this position, but the main motivation for Quine—and other philosophers working under his influence—is its power as a tool for revisionist ontology. Whereas it might seem like, say, fictional entities such as Pegasus exist in some sense—simply by virtue of the fact that we can talk about them—Quine’s account gives us an elegant way to avoid the weird and “unlovely” consequence of admitting their existence.11 To do so, the revisionist need only implement Russell’s doctrine of descriptions, translating sentences such as “Pegasus has wings” into the more obviously false sentence: “There is something that is Pegasus and has wings.” To the extent that we can recognize this sentence as obviously false, we can also have semantically perspicuous recourse to revisionism in ontology. In short, as Quine puts it, “the aesthetic sense of us who have a taste for desert landscapes” (1961: 4) is preserved by this brand of univocalism.

But if Quinean univocalism as presented in “On What There Is” is a preservation of some sort of aesthetic sense, then perhaps it is Peter Van Inwagen who has done the most to develop the case for univocalism in a more consciously argumentative register. In an influential essay, Van Inwagen (2009a) includes univocalism as one of five core theses which, together, constitute his broadly Quinean doctrine of ontological commitment.

Van Inwagen’s argument for univocalism can be summarized as a syllogism:

  1. Numerical terms are univocal in meaning.

  2. “Exists” is a numerical term.


  3. “Exists” is univocal in meaning.12

Van Inwagen treats (1) as axiomatic. “No one, I hope,” writes Van Inwagen, “supposes that number-words like ‘six’ or ‘forty-three’ mean different things when they are used to count objects of different sorts. The essence of the applicability of arithmetic is that numbers can count anything, things of any kind” (2009a: 482). Successful arithmetic reasoning would be impossible, suggests Van Inwagen, were numerical terms such as’6’ and’43’ equivocal as applied to objects residing in different ontological categories (such as, e.g., social entities and natural objects, respectively). Therefore, because successful arithmetic reasoning clearly is possible, numerical terms must not be equivocal in meaning. As Van Inwagen puts it, “[I]f you have written thirteen epics and I own thirteen cats, the number of your epics is the number of my cats” (2009a: 482).

The minor premise, (2), is perhaps not as intuitive as the first, but it is a staple of a tradition. Long before Quine and Van Inwagen, it was Frege who wrote, “Affirmation of existence is in fact nothing but denial of the number zero.”13 Van Inwagen concurs: “To say that unicorns do not exist is to say something very much like this: the number of unicorns is 0; to say that horses exist is to say essentially this: the number of horses is 1 or more” (2009a: 482). It is this “intimate connection” between “exists” and numericity that Van Inwagen leverages to justify the middle term of his argument.

The main advantages of this “number view” of existence are twofold. First, the Fregean (and Quinean) formula seems to offer a seamless, clear explication of the existence predicate.14 Given the otherwise radical lack of consensus in the history of Western philosophy about the meaning(s) of “exists,”15 any view that can rattle off an unambiguous answer to the question should be accorded merit. Second, according to Van Inwagen, there is nothing valuable that is “lost” in this rendering of the predicate—nothing, that is, that would hinder an otherwise adequate account of ontological commitment.

Univocalism about existence is motivated accordingly, then—both by its able service unto Quine’s “taste for desert landscapes,” and as the conclusion of Van Inwagen’s argument from numerical terms.

1.2. Motivations for the Aristotelian Alternative

Given that univocalism and the Aristotelian alternative are contradictory opposites, it should come as no surprise that their motivations are opposites as well, in some sense. One particularly influential historical motivation for the position that there are various modes of existence comes from Aristotle himself in Metaphysics Γ.2. The basic idea—one which Aristotle expects us to recognize intuitively—is that “accidents” seem to exist in different ways in relation to “substances,” that is, things with a more or less independent existence.16

So, for example, we say, “Socrates [a substance] is wise [an accident],” and it seems clear enough that both Socrates and Socrates’ wisdom exist. After all, neither Socrates nor his wisdom is nothing. Yet it would be strange to say that Socrates and his wisdom exist in the same way, since Socrates’ wisdom exists in an utterly dependent way, whereas Socrates himself exists in a relatively independent way.17 Presumably, Socrates would have gone on existing had he successfully refuted the Oracle at Delphi and shown himself to be a fool. By contrast, Socrates’ wisdom would not have existed without Socrates, except perhaps as a possibility. In this way, the Aristotelian analysis yields a distinction between two “modes” of existence, substantial and accidental. There are other modes of existence, but the point here is simply that such a distinction is made in order to account for the data of experience in a way that is less crude than merely attributing a single sense of existence to everything about which it is possible to speak meaningfully.18

Yet, like the Quinean “aesthetic” univocalism outlined above, the Aristotelian position as outlined in Metaphysics Γ.2 is more of a theoretical disposition than an argument. For a more dialectically adequate motivation for the Aristotelian alternative—one that meets Van Inwagen’s argument directly—we turn to Kris McDaniel’s recent work on “ways of being.”

McDaniel’s work fits our purposes here because he explicitly rejects univocalism while accepting the premises of Van Inwagen’s argument: namely, the judgment that “exists” is (among other things) a numerical term.19 Indeed, like Frege, Quine and Van Inwagen (among others), McDaniel accepts a quantificational framework for the semantics of existence.20 In fact, what McDaniel rejects in Van Inwagen’s univocalism is not so much the argument itself, but instead its exclusionary implications. In other words, according to McDaniel, while there is a (trivial) sense in which “exists” is univocal, there is also a sense in which it is not. The key move for McDaniel is his introduction of “naturalness” as an additional consideration for the discussion.

McDaniel introduces his notion of naturalness by contrasting it with a “mere disjunction”:

Consider the property of having a charge of −1 and the property of either being loved by Sarah Jessica Parker or having a charge of −1. Eddie the electron exemplifies both features. -1 charge is a real respect of similarity between electrons, but it is bizarre to think that Matthew Broderick and Eddie are similar in virtue of both of them enjoying either being loved by Sarah Jessica Parker or having a charge of −1. There is a metaphysical distinction between these two features: the former property carves nature at the joints, while the latter is a mere disjunction. (McDaniel 2017: 27, italics mine, for clarity’s sake)

Because mere disjunctions such as ‘has −1 charge or is loved by Sarah Jessica Parker’ are only trivially true when predicated of electrons, they are not natural in the way that simply ‘has −1 charge’ is. As McDaniel remarks, this is because the latter property is “a real respect of similarity” between electrons, whereas the mere disjunction is not. Although in a loose sense both “features” are truly predicated of the electron, only one carves nature at the joints.

On McDaniel’s analysis, then, the question of whether we should be outright univocalists or Aristotelians regarding existence is not a question about whether there is, in fact, a strictly numerical, univocal sense of the existential quantifier of standard predicate logic (∃). Indeed, McDaniel concurs that there is (2017: 31). Rather, the question is about the respective naturalness of (a) existence in this general, numerical sense (∃); and (b) instances of existence in different categories.21

If existence in general (∃) is perfectly natural, then we should be univocalists like Van Inwagen. This is because a single sense of “exists”—in the aforementioned strictly numerical sense—“carves nature at its joints” when it is truly predicated of anything at all. However, if existence in general is less natural than one or more of its instances—as a mere disjunction is less natural than one or more of its disjuncts—then the case for exclusionary univocalism suffers accordingly. After all, on the Aristotelian view, modes of existence are what carve nature at its joints, not existence in general.

It is precisely this latter position that McDaniel adopts, and he uses a fundamental distinction between material objects and Platonic abstracta to illustrate the case. Assuming that both material objects and Platonic abstracta exist in some sense, the univocalist must say that both kinds of object are simply instances of exactly one (perfectly natural) feature of existence.22 But this is potentially problematic, as McDaniel points out:

Material objects are necessarily temporal. It is hard to see what could ground this necessity if the location relation is metaphysically distinct from existence. On [univocalism], there are material objects and there are times, and there is a metaphysically primitive relation linking the two. [But] it is not at all clear why any material object must bear this relation to some time or other. (McDaniel 2017: 61)

In other words, while univocalism has no difficulty in stating that all material objects are, in fact, temporal, it does have significant difficulty in stating why they are so necessarily. The same point applies to Platonic abstracta: while univocalism can state that, in fact, these objects are a-temporal, it is not clear why this is case necessarily.

By contrast, if we accept that there are modes of existence that are more natural than existence in general, a more obvious answer is available: namely, that it is part of the very existence of material objects to exist temporally. That is to say, material beings are necessarily temporal because to exist as a material being just is (in part) to exist temporally.23 Accounting for this necessity in a straightforward manner is an explanatory advantage of the Aristotelian view, according to McDaniel. Since modes of existence are natural—not existence in general—there is no harm in including features such as temporality as “part of the package” of a material thing’s very existence. Indeed, if there are modes of existence, this is exactly what we should expect when it comes to accounting for existing things that are so fundamentally different. Without offense to the ordinary language user, the Aristotelian can simply say things like “material objects exist in time,” whereas Platonic abstracta do not.

Finally, the Aristotelian view enjoys these advantages all while preserving the abovementioned, basic motivation for univocalism: namely, that existence is importantly analogous to number, and that normal arithmetic reasoning requires that we use numbers univocally. Indeed, the Aristotelian is quite happy to admit this: arithmetic reasoning is both univocal and possible. Crucially, however, what the Aristotelian sees is that, qua arithmetician, one prescinds from a more properly metaphysical concern of naturalness. Therefore, at least to the extent that we admit that some properties are more natural than others, we must go beyond arithmetic reasoning. In short, we must do metaphysics.

2. Univocalism and the Aristotelian Alternative: An Intuition Pump

Needless to say, the motivations for these two accounts of existence barely rise to the level of snapshots. The problems could of course be analyzed in other ways. But again, my urgent task is to demonstrate a connection between (a) questions about existence in general to (b) questions of social ontology. Here is one way to think about it at the level of basic intuitions.24

The following puzzle is an “aporetic tetrad,” that is, a puzzle consisting of four apparently plausible propositions that seem inconsistent when taken together, and for whom the affirmation of any three entails the negation of the remaining one. In effect, the puzzle amounts to an exercise in testing our intuitions about which three propositions are true, and which one is false.

Consider a social group (e.g., a nation-state) materially constituted by a finite number of individual people and things (and nothing more):

  1. The social group exists.

  2. The individual people/things exist.

  3. The social group is not materially distinct from the individual people/ things constituting it.

  4. “Exists” in (1) and (2) is univocal, that is, they both signify exactly one mode of existence.

The point of this puzzle is to illustrate the potential relevance of one’s doctrine of existence for questions about integral wholes in general—and for social group, in particular—insofar as many non-reductivist approaches to social groups rely upon the intuitive sense that some social wholes are conceptually and/or ontologically irreducible to their parts/members—usually on some account of novel causal powers that are present in the whole but not in the parts. Assuming that proposition (3) can be affirmed by univocalists and Aristotelians alike, I submit that the univocalist’s “natural” intuition is to reject (1). Because there is only one way in which a thing might exist, there is “competition” between the terms of what otherwise might appear as a parts-to-whole relation.25 The univocalist is presented with a choice: either the “parts” or the “whole” have to go, and given the parts’ relative independence with respect to the whole, it only makes sense to do away with the latter.

By rejecting (4), however, the Aristotelian can accommodate both the social group and the individual people/things, since she can simply say that the whole’s mode of existence is distinct from the parts’ mode of existence.26 Indeed, the situation resonates deeply with the abovementioned, paradigmatic cases of the Aristotelian doctrine of substance and accident, as well as McDaniel’s doctrine of the merely analogical similarity enjoyed by “existing” abstracta and temporal objects, respectively. That is to say, for reasons explored above by McDaniel, the distinction between the social group and the individual people/things seems to demand consideration at the level of modes of existence, as opposed to a more conventional consideration of standard first-order properties.

3. Tuomela’s “Fictionalism”: A Case Study

Having cleared some ground at the level of general metaphysical intuitions, I will now try “put some flesh” on the discussion by examining a recent debate between Raimo Tuomela and Michael Schmitz on the nature of social groups and their existence. Again, the goal is to see what difference the Aristotelian position on existence might make for an unapologetically realist account of social groups.

This debate concerns the account of social groups given in Tuomela’s most recent monograph, Social Ontology, which provides a theoretical account of the fundamental thesis that “social groups, including large organized groups, can be viewed as functional group agents. This means that we can on functional [and conceptual] grounds attribute as-if mental states such as wants, intentions, and beliefs, as well as actions and responsibility to these groups” (2013: x). In other words, according to Tuomela, when we say things like “the United States is at war,” there is at least one sense in which we are not simply speaking in “shorthand” about some large aggregate of individual people and things who are “really” the ones doing the relevant being, intending and acting.27 Rather, we are speaking about a social group: namely, the United States, and its own, proper intentions and activities. This claim has both functional and conceptual justifications, for Tuomela.

3.1. Justifying the Irreducibility of Social Group

It is easy to see the claim’s functional justification, since it is obviously impossible to account for every relevant fact about all the individual people/things constituting the United States. Thus, to the extent that sentences such as “the United States is at war” are intelligently affirmed or denied, we have to make use of what is at least a “rough and ready” concept of the United States as such, that is, without recourse to the psychology of all relevant individuals.

More interesting for our purposes, however, is Tuomela’s conceptual justification, which he establishes via his account of “we-mode” collective intentionality. According to Tuomela, “to think … and act in the we-mode is to think and act fully as a group member. This represents a mode of thinking and acting, to act we-modely” (2013: 37). Without getting into the rich details of this thesis, for now it is enough to say that Tuomela understands social groups to be constituted by individuals enjoying we-mode collective intentionality. Thus, to the extent that we-mode intentionality (i.e., intending qua group member) is conceptually irreducible to the I-mode (i.e., intending qua individual), so are social groups conceptually irreducible to the individual people and things of which they are comprised.

Therefore at least one crucial part of Tuomela’s project is to identify the sense in which we-mode intentionality is irreducible to the I-mode. He does so by making recourse to a standard game-theoretic framework: “reducibility [of we-mode to I-mode] fails because we-mode reasoning leads to a set of action equilibria different from what individualist, I-mode theorizing leads to” (2013: 11). That is, when we-mode intentionality is contrasted with its I-mode counterpart in familiar game-theoretic scenarios, different solutions result. Tuomela’s example is this “Hi-Lo game (without communication):28

 Hi  Lo
 Hi  3,3  0,0
 Lo  0,0  1,1

Whereas purely I-mode or individualistic reasons cannot recommend what should be the obviously best solution of Hi-Hi, the point is that the group reasoning proper to we-mode intentionality (i.e., intending/acting as and for the good of the group) can.29 Thus, to the extent that we have different solutions in the we-mode than in the I-mode, it seems that the former cannot be reduced to the latter. Intending and acting in the we-mode makes a real difference when compared to the I-mode.

This is what Tuomela means when he says that groups intend and act: no more—and no less—than that suitably related individuals intend and act together in the we-mode.30

3.2. Tuomela’s “Fictionalism”

To recap, Tuomela argues that social groups “can be said” (2013: 32) to intend and act in a way that is irreducible to the intentions and actions of the people comprising them. After all, groups are formed and constituted by individuals intending and acting in the we-mode, and that the we-mode is conceptually irreducible to the I-mode is manifest in classic game theoretic scenarios. Further, this position is conceptually justified: we-mode intentionality makes a causal difference, as we have seen.

Given these commitments, then, it seems clear that Tuomela’s account of group agency should yield the correlative judgment that groups as such exist, that is, in a way that is irreducible to the existence of the individuals comprising them. After all, it seems that existence is a condition for agency.

Yet, as Michael Schmitz has pointed out, this isn’t so clear—at least not according to Tuomela himself. In fact, Tuomela says that group agents are “fictitious” on account of their having “fictitious features” (2013: 47). He points to two passages from Social Ontology, in particular, in order to demonstrate this ambiguity:

What does it mean to say that a group agent is fictitious and has fictitious features? My view is that group agents are mind-dependent entities and fictitious in the mind-dependence sense that involves collective imagination, idealization, and construction. They do not exist as fully intentional agents except perhaps in the minds of people (especially group members). This also makes the intentional states attributed to them fictitious because the bearers (viz., group agents) of these states are fictitious (not real except in the minds of the group members). (2013: 47)

Only the intentional properties attributed to groups are fictitious in the mere mind-dependence sense. Group agents qua nonintentional systems have causal powers and are capable of causing outcomes in the real world. (2013: 47)

Indeed, in other passages that Schmitz does not cite, Tuomela also says that we-mode groups can be said to intend and act “somewhat metaphorically,” (2013: 70) which of course calls into question whether or not it is literally true that, for example, the United States is at war. Indeed, he admits as much quite clearly.31

What Schmitz takes issue with—rightly, in my view—is Tuomela’s “marriage of fictionalism and causal realism” (Schmitz 2017: 43). If it is literally true that we-mode groups and I-mode individuals yield different action equilibria, but literally false to say that we-mode groups intend and act, then the only way to save the coherence of the position is to infer that the real causal efficacy of we-mode groups is non-intentional.32 This is an odd result, of course, if only because a we-mode group is supposed to be constituted by, well, we-mode, collective intentionality.33 Further, as Schmitz notes, Tuomela seems to be abusing the term “fictional.” After all, “real corporations like IBM are not fictitious in the sense in which Ewing Oil is fictitious” (Schmitz 2017: 47).

I cannot help but agree with Schmitz here: fictionalism and causal realism make for quite the odd couple!

4. The Aristotelian Alternative Put to Work

Now, as far as I know, Tuomela has never commented one way or another regarding the doctrine of modes of existence. Yet I cannot help but notice the way in which his “fictionalism” follows the reductionist pattern outlined in the aporetic tetrad above. If only implicitly, his distinction between “intrinsic” and “extrinsic” intentionality—had by individuals and groups, respectively—betrays a latent commitment to univocalism. After all, what appears to be a relation between affirmations, that is, that individuals and groups are intrinsic and extrinsic agents, respectively, actually turns out to be an affirmation and a negation: there are individual agents, but there are not (literally) group agents.

Yet, as we have seen above, these sorts of situations are quite different when viewed against the horizon of the Aristotelian position. That is to say, if individual persons and group agents exist in different ways, there is simply no temptation towards this marriage of causal realism and fictionalism. This is because, on the Aristotelian position, modes of existence are what “carve nature at its joints”—not existence in general (∃). There is no threat of paradox when it comes to affirming the existence of both individuals and the groups they comprise, since the two entities are not “competing” for the same ontological “space.”

In fact, the doctrine of modes of existence may even be better at accommodating Tuomela’s distinction on his own terms, since to distinguish intrinsic and extrinsic modes of intentionality and agency seems precisely to distinguish between two modes of having intentionality. Indeed, despite the aforementioned passages from Social Ontology, Tuomela seems explicitly to adopt something like this interpretation of his intrinsic-extrinsic distinction in his reply to Schmitz.34 After all, no one doubts that such social groups are “mind-dependent” in some sense. What is at issue is whether something can be both causally efficacious and fictional (i.e., non-existent).

I submit that Tuomela’s “marriage” of causal realism and fictionalism is problematic. The better path for Tuomela’s otherwise rigorously anti-reductionist approach to social groups is to reject this implicit univocalism and embrace a metaphysics that admits modes of existence. Indeed, such a view follows the same pattern—and enjoys the same advantages—as the one outlined by McDaniel with respect to abstracta and material objects. Just as material objects are necessarily temporal by virtue of their unique mode of existence, so are (paradigmatic) social groups necessarily mind-dependent (or “ontologically subjective”35) by virtue of their unique mode of existence.

Once again, the key metaphysical commitment here is that the general, numerical quantifier (∃)—and therefore existence in general—is less natural than its modes. Once this is admitted, one need not fret about the “ontological competition” outlined above. There is a sense in which the Aristotelian position allows us to be more faithfully empirical when it comes to our ontological commitments. This is because one’s recognition of causal agency in the world does not need to be “hemmed in” by preconceived, reductionist tendencies at the level of general metaphysics. In other words, we need not be embarrassed: if it is an agent, then it exists.36

5. Objections and Replies

Before concluding, I consider two key objections to the above argument.37

5.1. Is Univocalism Revisionary Per Se?

First, it might not be clear that univocalism about existence is revisionist/ reductionist per se. Even if it is true historically that influential univocalists happen to have shared a preference for desert landscapes, might this be accidental to univocalism as a position? After all, this objector might continue, can we not point to other advocates of univocalism, such as Frege, Van Inwagen, and even Quine himself—all of whom were/are committed to the existence of (e.g.) Platonic abstracta of one form or another?38

In response to this objection, I should say first that I agree with William Vallicella that, whatever their differences, all these versions of univocalism face serious difficulties when it comes to coherently affirming the existence of anything that is not plausibly interpreted as a “value of a bound variable.” This includes of course such things as concepts, which are supposed to be instantiated by such “values,” and individual constants, which by definition cannot be values of bound variables.39 This is no surprise to the Aristotelian, since from her perspective such a result is precisely what we should expect from one who has failed to appreciate the properly (ontological) categorical differences between entities.40 Therefore, especially if social groups turn out to be categorically different from other entities, as some have argued,41 there is reason to think that univocalism “tends towards” something like individualist reductionism.

But even if the univocalist could come up with satisfactory re-descriptions that would neutralize these sorts of difficulties, there is an even more fundamental methodological distinction at work, as Jonathan Schaffer has noted: namely, the distinction between the inherently “revisionary” character of the Quinean position, as opposed to “permissive” Aristotelian position (Schaffer 2009: 353). According to Schaffer, while both positions are constrained by parsimony with respect to ontological commitment, this is the case in different ways.

For the Quinean—and for the univocalist more generally, I suggest42—the fundamental concern of metaphysics is to posit exactly as many entities as there (really) are. In Schaffer’s words, “The task is to solve for E = the set … of entities” (2009: 354). On this model of ontological commitment, to be constrained by parsimony is to go about this task with a default preference against admitting social groups into the one category that really matters: namely, E. If there are reasons to deem such (alleged) entities dubious or superfluous, then ceteris paribus the parsimonious Quinean is obliged to adopt an anti-realist position.

For the Aristotelian, by contrast, obedience to parsimony does not imply this default suspicion of entities with respect to their (generic) existence. On the contrary, the Aristotelian can be quite permissive in this respect.43 This is because parsimony for the Aristotelian constrains not the “roster of existents” in general, but rather the “grounds” or (in McDaniel’s terms) “purely natural” entities that lie at the foundation of that ordered roster (Schaffer 2009: 353). In other words, for the Aristotelian, the fundamental question of group realism is not whether social groups exist (“Of course they do!” she replies), but rather how they exist (i.e., to what extent they are natural). It is precisely this permissiveness that offers the Aristotelian a clearer and more direct path to group realism than is available to the revisionary univocalist.

5.2. Supervenience?

The second objection concerns the abovementioned aporetic tetrad. It might not be so clear that a realism that coherently affirms the existence of both individuals and groups requires the admission of distinct modes of existence in order to do so. Could not a Quine-inspired supervenience account do the job of adequately accounting for social groups, that is, as “entailed” or “necessitated” by facts about individuals?44 On such a view, the four limbs of the tetrad can all be affirmed, since proposition (1)—namely, that the social group exists—is glossed in such a way that all facts about the social group supervene on facts about relevant individuals. This is the approach taken by Christian List and Philip Pettit, who carve out a sensible “moderate” position on social groups that is “realist” but not “emergentist” (List & Pettit 2011: 9). According to List and Pettit, facts about individual attitudes and intentions logically necessitate facts about group attitudes and intentions in the way that suitably arranged dots on a grid necessitate the shapes they materialize, that is, as one set of facts “fixes” another (2011: 65). When we understand this (weaker) sense in which social groups exist (i.e., as supervenient on facts about individuals), then the tetrad ceases to be aporetic, and we can hang on to univocalism.

Now of course it is impossible for me to do justice to List and Pettit’s complicated and well-developed position here. Still, for my purposes, it is worth mentioning that List and Pettit’s “realism” about social groups bears a striking resemblance to Tuomela’s in one very important sense: namely, that it is hardly realism at all. Indeed, as List and Pettit themselves say, “The autonomy we ascribe to group agents under our approach is epistemological rather than ontological in character” (2011: 76). Facts about social groups supervene on those of individuals, and so the social group is not an addition to the roster of existents.

If anything, List and Pettit’s account of group agency is even less realist than Tuomela’s, since the autonomy they grant unto social groups seems to be based in the functional (as opposed to conceptual) impossibility of accounting for facts about social groups in terms of those about individuals.45 All in all, it is difficult to see how the supervenience account given by List and Pettit can credibly affirm proposition (1) of the tetrad. As D. M. Armstrong remarks pithily, “[W]hatever supervenes … is not something ontologically additional to the subvenient” (1997: 12).

Therefore, the truth or falsity of List and Pettit’s supervenience account of social groups aside,46 I suggest that it is simply not the realism that critical social ontologists ought to be interested in.47

6. Conclusion

Critical social ontologists are realists when it comes to the existence and causal efficacy of social groups. They think that social groups do things in the world. What I have argued in this paper is that philosophers who might otherwise share these realist inclinations have something to gain from the Aristotelian doctrine of modes of existence. In short, there is simply no need to “hedge” our ontological commitments, as Tuomela does. Individual persons and social groups exercise intentionality/agency in different ways because they exist in different ways. But this implies that indeed there are modes of existence. The two positions are correlatively linked: to the extent that the Aristotelian position is plausible, so is the realist account of social groups for which critical social ontologists are advocates.


  1. See, e.g., Horkheimer (2014).
  2. See, e.g., Groff (2013).
  3. See Haslanger (2003), Witt (2011), Barnes (2017), Isaacs (2011), Átsa (2018), Sundstrom (2002), and Root (2000).
  4. This is the substantive question of the contemporary debate between so-called “thin” (univocalist) and “thick” (Aristotelian) theorists of existence. For an overview of this dispute, see Vallicella (2014: 46–48).
  5. This debate can be found in Schmitz (2017) and Tuomela (2017).
  6. I’m using the term “clearing” in the sense of Heidegger’s Lichtung—that is, a condition of possibility for the appearance of distinct objects in a world. See Heidegger (1986: 380).
  7. The question of what ontological categories are is a difficult one. For my purposes, it is enough to say that they are distinct genera of being, where “genera” is taken in a technical sense, i.e., as an internally differentiated kind. See Aristotle (1967: 1025a).
  8. This is not to say, however, that there are not versions of univocalism that are compatible with the Aristotelian thesis. On this point, see discussion below.
  9. For this view, see Ritchie (2020).
  10. See, e.g., Frege (2001).
  11. As Quine puts it, an “Wyman’s overpopulated universe is in many ways unlovely… .[W] e’d do better simply to clear Wyman’s slum and be done with it” (1961: 4).
  12. See Van Inwagen (2009a: 482).
  13. Quoted in Van Inwagen (2009a: 483).
  14. It is important that Quine does not purport to offer a definition of the existence predicate; rather, he offers an explication. An explication, in Quine’s terms, is more than a “paraphrase [of the] definiendum into an outright synonym,” but it still relies upon “other pre-existing synonymies” for its meaningfulness. On this point, see Quine (1961: 25).
  15. For an accessible introduction to the radical diversity of views on the existence predicate in the history of philosophy, see Gilson (1952).
  16. Aristotle puts the point as follows in Metaphysics Γ.2: “For some things are said to “be” because they are substances; others because they are modifications of substance; others because they are a process towards substance, or destructions or privations or qualities of substance, or productive or generative of substance or of terms relating to substance, or negations of certain of these terms or of substance” (1967: 1003b).
  17. One might object that it could simply be a property of the nature of wisdom to be profoundly dependent upon a substance, and therefore that there is no need to invoke modes of existence in order to account for this modal asymmetry between substance and accident. For a reply to this sort of objection, see the discussion below.
  18. The point here is that Aristotle takes his position on existence to be justified to the extent that it allows the enterprise of ontological commitment to be faithfully “empirical” even in the face of analytical difficulties. As he remarks in Metaphysics θ, “[W]e should not seek a definition of everything, but should also perceive [some things] by analogy” (1967: 1048a).
  19. See McDaniel (2017: 22).
  20. That is to say, he accepts that the meaning of “exists” in ordinary language can be faithfully represented by the so-called existential quantifier of standard predicate logic (∃).
  21. This is especially so for instances of existence that are so fundamentally dissimilar that they invite consideration (by Aristotelians, at least) of distinct ways or modes of existence.
  22. I use the generic term “feature” here rather than “property” because, on McDaniel’s view, there is no need to commit oneself to the view that existence is a first or second-order property, or even a property at all. On this point, see McDaniel (2017: 56).
  23. As McDaniel puts it, “[T]here is no mystery if what it is for a material object to be [i.e., to exist] is for it to be at some time. It is part of the very being of a material being that it is in time” (2017: 61).
  24. I borrow the form of the following “puzzle” from Vallicella (2014: 67–68).
  25. It is perhaps worth noting that John Searle levels this sort of objection against the very project of theorizing social entities such as groups in the first place. Instead of social entities, Searle suggests that the proper subject matter of social ontology ought to be social (or institutional) facts. For a brief overview of the debate about what the proper subject matter of social ontology is (or ought to be), see Epstein (2018).
  26. This analysis is derived from Vallicella (2014: 67–68).
  27. As we will see, whether we can do so literally is unclear, on Tuomela’s analysis.
  28. For the entire context of Tuomela’s argument about game-theoretical implications of his account of the we-mode, see Tuomela (2013: 179–213).
  29. We should be careful here: it is not that classical game theory cannot offer any recommendations at all when it comes to such a Hi-Lo game; rather, as Tuomela puts it, “[W]e-mode reasoning can help to give an explanation of human cooperative behavior in collective action dilemmas that is left unexplained by standard noncooperative game theory” (2013: 180).
  30. Indeed, according to Tuomela, “[We-mode] intentionality can be regarded as ‘the cement of society’ ” (2013: x).
  31. “That a group’s intention or belief, etc., is fictitious entails that it is not literally true that it intends or believes, etc.” (Tuomela 2013: 47).
  32. “However, the price to be paid for this marriage of fictionalism and causal realism about group agents is that the relevant causation is not intentional causation” (Schmitz 2017: 43).
  33. In Tuomela’s defense, his position is that groups have “extrinsic, derived” intentionality, as opposed to “intrinsic” intentionality. That this makes the intentional properties of groups “fictional,” however, seems to me a strange inference. For Tuomela’s response to Schmitz on this point, see Tuomela (2017: 71–78).
  34. “Note that my view does not take group agents, such as we-mode groups, to be more than partly fictitious because group agents are still real in a causal sense, as said in so many words in SO” (Tuomela 2017: 72).
  35. This term comes from Searle, and while he does not give a robust account of the matter, he does admit modes of existence to account for the distinction between ontologically subjective and ontologically objective entities. See Searle (1998: 44).
  36. For precisely this reason, I submit that the doctrine of modes of existence is intimately connected—and perhaps even necessary for—the critical realist Roy Bhaskar’s “causal criterion of reality.” For this concept in Bhaskar’s work, see Bhaskar (1998: 50).
  37. Thanks to two anonymous referees for leveling these objections.
  38. See, respectively, Frege (2001), Van Inwagen (2009b), and Quine (1951).
  39. That is, unless they are values of a haecceity concepts. For an argument against haecceity concepts/properties, see Vallicella (2002: 99–104).
  40. Again, for the Aristotelian, entities belonging to different categories have different modes of existence because the categories just are these modes.
  41. See Clarke (2009: 39–47).
  42. I should note that Schaffer himself appears to espouse univocalism of a sort: “I am invoking the one and only sense of existence, and merely holding that very much exists” (Schaffer, 360). But such a remark is perfectly compatible with the Aristotelian view if what is meant by this one “sense” of existence is the generic quantifier, as elaborated by McDaniel.
  43. As Schaffer notes (correctly, in my view), “While Quine is interested in existence questions (such as whether there are numbers), Aristotle seems to take a permissive disinterest in such questions. … He simply assumes that all such types of entity exist [i.e., in a generic sense], without need for further discussion” (2009: 352).
  44. For this view of supervenience, see Armstrong (1997: 12).
  45. According to List and Pettit, “While the agency achieved by a group supervenes on the contributions of its members—while it is not ontologically autonomous—it is autonomous in another, related sense. The agency of the group relates in such a complex way to the agency of individuals that we have little chance of tracking the dispositions of the group agent, and of interacting with it as an agent to contest or interrogate, persuade or coerce, if we conceptualize its doings at the individual level” (2011: 76).
  46. For some notable criticisms of supervenience accounts of social groups, see Epstein (2009) and (2015).
  47. Perhaps not coincidentally List and Pettit appear to reject the notion that “the market” contains socially emergent properties as a structured whole, for example (e.g., List & Pettit 2011: 12)—hardly a result that a critical social ontologist should countenance.


Aristotle (1967). Aristotelis Metaphysica. Werner Jaeger (Ed.). Oxford University Press.

Armstrong, D. M. (1997). A World of States-of-Affairs. Cambridge University Press.

Átsa (2018). Categories We Live By: The Construction of Sex, Gender, Race, & Other Social Categories. Oxford University Press.

Barnes, Elizabeth (2017). Realism and Social Structure. Philosophical Studies, 10, 2417–33.

Bhaskar, Roy (1998). The Possibility of Naturalism: A Critique of the Contemporary Human Sciences. Routledge.

Clarke, William Norris (2009). System: A New Category of Being? In The Creative Retrieval of St. Thomas Aquinas: Essays in Thomistic Philosophy, New and Old (39–47). Fordham University Press.

Epstein, Brian (2009). Ontological Individualism Reconsidered. Synthese, 166(1), 187–213.

Epstein, Brian (2015). The Ant Trap: Rebuilding the Foundations of the Social Sciences. Oxford University Press.

Epstein, Brian (2018). Social Ontology. In Edward N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy.

Frege, Gottlob (2001). Dialog mit Pünjer über Existenz. In Gottfried Gabriel (Ed.), Schriften zur Logik und Sprachphilosophie (1–22). Felix Meiner Verlag.

Gilson, Etienne (1952). Being and Some Philosophers. Pontifical Institute of Medieaval Studies.

Groff, Ruth (2013) Ontology Revisited: Metaphysics in Social and Political Philosophy. Routledge.

Haslanger, Sally (2003). Social Construction: The ‘Debunking’ Project. In Frederick F. Schmitt (Ed.), Socializing Metaphysics: The Nature of Social Reality (301–25). Rowman & Littlefield.

Heidegger, Martin (1986). Seminar in Zähringen. In Gesamtausgabe 15 (372–400). Vittorio Klostermann.

Horkheimer, Max (2014). Traditional and Critical Theory. In Ruth Groff (Ed.), Subject and Object: Frankfurt School Writings on Epistemology, Ontology, and Method (185–232). Bloomsbury Academic.

Isaacs, Tracy (2011). What Would a Feminist Theory of Collective Action and Responsibility Look Like? In Kendy Hess, Violetta Igneski, and Tracy Issacs (Eds.), Collectivity: Ontology, Ethics, and Social Justice (223–40). Rowman & Littlefield.

List, Christian and Philip Pettit (2011). Group Agency: The Possibility, Design, and Status of Corporate Agents. Oxford University Press.

McDaniel, Kris (2017). The Fragmentation of Being. Oxford University Press.

Quine, W. V. O. (1961). From a Logical Point of View. Harper and Row.

Renault, Emmanuel (2016). Critical Theory and Processual Social Ontology. Journal of Social Ontology, 2(1), 17–32.

Ritchie, Katherine (2015). The Metaphysics of Social Groups. Philosophy Compass, 10(5), 257–72.

Ritchie, Katherine (2020). Social Structures and the Ontology of Social Groups. Philosophy and Phenomenological Research, 100(1), 402–24.

Root, Michael (2000). How We Divide the World. Philosophy of Science, 67(3), 628–39.

Schaffer, Jonathan (2009). On What Grounds What. In David J. Chalmers, David Manley and Ryan Wasserman (Eds.) Metametaphysics: New Essays on the Foundations of Ontology (347–83). Oxford University Press.

Schmitz, Michael (2017). What Is a Mode Account of Collective Intentionality? In Gerhard Preyer and Georg Peter (Eds.), Social Ontology and Collective Intentionality: Critical Essays on the Philosophy of Raimo Tuomela with His Responses (37–70). Springer.

Searle, John (1998). Mind, Language, and Society: Philosophy in the Real World. Basic Books.

Sundstrom, Ronald (2002). Racial Nominalism. Journal of Social Philosophy, 33(2), 193–210

Thompson, Michael J. (2017). Social Ontology and Social Critique: Toward a New Paradigm for Critical Theory. In Daniel Krier and Mark P. Worrell (Eds.), The Social Ontology of Capitalism (15–45). Palgrave Macmillan.

Tuomela, Raimo (2013). Social Ontology: Collective Intentionality and Group Agents. Oxford University Press.

Tuomela, Raimo (2017). Response to Michael Schmitz. In Gerhard Preyer and Georg Peter (Eds.), Social Ontology and Collective Intentionality: Critical Essays on the Philosophy of Raimo Tuomela with His Responses (71–78). Springer.

Vallicella, William F. (2002). A Paradigm Theory of Existence: Onto-Theology Vindicated. Springer.

Vallicella, William F. (2014). Existence: Two Dogmas of Analysis. In Daniel D. Novotný and Lukáš Novák (Eds.), Neo-Aristotelian Perspectives in Metaphysics (45–75). Routledge.

Van Inwagen, Peter (2009a). Being, Existence, and Ontological Commitment. In David Chalmers, David Manley, and Ryan Wasserman (Eds.), Metametaphysics: New Essays on the Foundations of Ontology (472–506). Oxford University Press.

Van Inwagen, Peter (2009b). God and Other Uncreated Things. In Kevin Timpe (Ed.), Metaphysics and God: Essays in Honor of Eleanore Stump (3–20). Routledge.

Witt, Charlotte (2011). The Metaphysics of Gender. Oxford University Press.