The view that social kinds (e.g., money, migrant, marriage) are mind-dependent is a prominent one in the social ontology literature.1 However, in addition to the claim that social kinds are mind-dependent, it is often asserted that social kinds are not real because they are mind-dependent.2 For instance, F.A. Hayek claims that social kinds are not defined by “real properties” because social kinds “are what people think they are” (Hayek 1943: 3). Similarly, John Searle argues that social kinds are not real because they “only are what they are, because that is what we believe that they are” (Searle 2007: 4). And Amie Thomasson argues that many social kinds are not real because they exist and have their nature in virtue of our thoughts about them (Thomasson 2003b).3 Call this view social kind anti-realism:
Social kinds are not real because they are mind-dependent.4
In what follows, I argue that social kind anti-realism is not well founded.5
To defend their view, social kind anti-realists must accomplish two tasks. First, they must show that social kinds are mind-dependent. That is, they must identify a dependence relation that obtains between social kinds and our mental states. Call this the Dependence Task. Second, they must show that social kinds are not real because of the dependence relation that obtains between them and our mental states. Call this the Anti-Realist Task.
A relation can accomplish the Anti-Realist Task only if it accomplishes the Dependence Task. However, accomplishing the Dependence Task does not thereby accomplish the Anti-Realist Task. In other words, demonstrating that social kinds are mind-dependent is not sufficient to demonstrate that they are not real. Social kind anti-realists must also show that being mind-dependent in the relevant sense makes social kinds unreal.6
The structure of my argument against social kinds anti-realism will proceed as follows. I consider several ways of accomplishing the Dependence Task. That is, I consider several different ways of defining the relation that is supposed to obtain between social kinds and our mental states. Some of the relations I consider are explicitly defended in the social ontology literature. Others are defended implicitly or are implied by explicitly defended views. With respect to each relation, I argue that either it fails to accomplish the Dependence Task, or it fails to accomplish the Anti-Realist Task. As such, anyone who wishes to defend social kind anti-realism must provide an alternative explanation of how social kinds depend on our mental states in a way that impugns their reality. In the absence of such an explanation, there is no reason to endorse social kind anti-realism.7
Philosophers use the word ‘real’ to mark a wide variety of distinctions. As Crispin Wright puts it:
If there ever was a consensus understanding about ‘realism,’ as a philosophical term of art, it has undoubtedly been fragmented by the pressures exerted by various debates—so much so that a philosopher who asserts that she is a realist about theoretical science, for example, or ethics, has probably, for most philosophical audiences, accomplished little more than clear her throat. (1992: 1)
Given this, it is especially important for proponents of social kind anti-realism to clearly state the sense in which they claim that social kinds are not real. In this section, I argue that social kind anti-realists have failed to do so. To demonstrate that this is the case, I consider three prominent ways of characterizing the realism/anti-realism distinction. I argue that none of these captures the sense in which social kinds fail to be real according to social kind anti-realists. The fact that social kind anti-realists have not identified the sense of unreality that they attribute to social kinds gives us a preliminary reason to doubt that social kind anti-realism is well-founded.
One prominent use of the term ‘real’ is to distinguish between existent and non-existent entities: real entities exist, unreal entities do not. But those who defend social kind anti-realism do not argue that social entities are non-existent.8 Rather, social kind anti-realists are what Gideon Rosen call ‘modern idealists’—that is, anti-realists who admit that the allegedly unreal entities in question exist (Rosen 1994: 289). According to social kind anti-realism, social kinds (e.g., money) exist but their ontological status is diminished in some way.
Another prominent use of the word ‘real’ corresponds to the idea of fundamentality. On this view, entities are real only if they are metaphysically fundamental (Fine 2001). However, like the whole of the social world, the vast majority of the natural world is not metaphysically fundamental either. Yet proponents of social kind anti-realism do not argue that such things as water and electrons are not real. Rather, they maintain that physical, chemical, biological, and even psychological kinds are real, but deny that social kinds have the same status.
Third, some philosophers deny that a domain of entities is real if our discourse about them is not truth apt. On this use of the term ‘real,’ if our assertions about some entities are not in the business of being true or false, then those entities are not real. But social kind anti-realists about social kinds do not claim that our assertions about them are not truth apt. According to proponents of social kind anti-realism, it is true, for example, that bills issued by the Bureau of Engraving and Printing are money, and it is false that blue gym socks are money.
Most often, the sense of reality that social kinds are supposed to lack is characterized figuratively. Indeed, the relevant figurative descriptions are common to many discussions of realism and anti-realism in the philosophical literature. Gideon Rosen describes the distinction between realism and anti-realism as follows:
What the realist mainly claims is the right to say things like this: “Our discourse about X concerns a domain of fact that is out there. These facts obtain anyway, regardless of what we may think. When all goes well, inquiry into the disputed area discovers what is already there, rather than constituting or constructing its object. Successful thought amounts to the detection of something real, as opposed to a projection onto the real of our own peculiar or subjective perspective…” And so on down the list of familiar words and pictures. The antirealist denies our right to this imagery. (1994: 278 emphasis in original)
Thus, what social kind anti-realists deny is that social kinds are ‘out there.’ They argue that facts about them do not obtain ‘anyway, regardless of what we may think.’ Rather, the way in which social kinds depend on our mental states is such that our minds ‘constitute’ or ‘construct’ them. Therefore, social kinds are merely a ‘projection’ of our thoughts onto mind-independent reality. Though these descriptions are certainly evocative, they do not add up to a clear definition of the way in which social kinds are supposed to be unreal. Thus, I maintain that social kind anti-realism is not a well-defined view.
Nonetheless, in subsequent sections I consider various relations that might obtain between social kinds and our mental states. I argue that none of these relations substantiates the anti-realist’s rhetoric. Of course, proceeding with a merely figurative characterization of the sense of unreality social kind anti-realists attribute to social kinds is unsatisfying. All else being equal, it would be better to criticize a view that is formulated in a clear and precise way. Nonetheless, there is no prohibition on criticizing unclear and poorly developed views. What we can do is point to the unclarity in a forthright way, earnestly try to clarify it, and, if that fails (as I have argued that it does), criticize social kind anti-realism in such a way that the unclarity does not need to be resolved in order to see why the view does not support even the figurative claims of unreality. This is what I intend to do in what follows.
Here is how the remainder of the paper will proceed. In section 3, I consider two ways of analyzing the way in which social kinds are supposed to depend on our mental states: a modal-existential relation and a ground-theoretic relation. I argue that the modal-existential relation fails to accomplish the Dependence Task and that the ground-theoretic relation fails to accomplish the Anti-Realist Task.
In sections 4 and 5, I consider the proposal that social kinds depend on our mental states in the sense that we determine which properties are essential to them. I argue that we cannot determine the identity or nature of any kind, social or not. As such, the proposed analysis of mind-dependence likewise fails to accomplish the Dependence Task.
Finally, in section 6, I consider Ásta’s suggestion that social kinds are mind-dependent in the sense that they are defined by conferred properties, i.e., properties that are instantiated in virtue of subjects’ attitudes toward the entities that instantiate them (2008, 2013, 2018). I argue that even if this relation can accomplish the Dependence Task, it fails to accomplish the Anti-Realist Task.9
Probably the most widely known and widely accepted version of the thesis that social kinds depend on our mental states is the collective acceptance view of social kinds. Different versions of the collective acceptance view have been defended by John Searle (1995, 2010), Amie Thomasson (2003a, 2003b), Raimo Tuomela (2007, 2013), and, less explicitly, by Muhammad Ali Khalidi (2015).10 According to the collective acceptance view, a social kind, K, depends on our mental states in the sense that K is instantiated only if some Xs satisfy some collectively accepted conditions, C, for being K. On this view, if some Xs satisfy C, then the Xs are K.
Though Searle is the most well-known proponent of the collective acceptance view, I will focus on Amie Thomasson’s formulation of the view because it has the virtue of being both clear and concise:
(MD1) A kind, K, is mind-dependent =df An entity, x, is K only if we collectively accept that some conditions, c1…cn, suffice for being K.11
For example, according to the collective acceptance view, something is money only if we collectively accept that some conditions (e.g., being a bill issued by the Bureau of Engraving and Printing) suffice for being money.12 If we collectively accept that being a bill issued by the Bureau of Engraving and Printing is sufficient for being money, then anything that is a bill issued by the Bureau of Engraving and Printing is money.
In what follows, I argue that MD1 does not accomplish the Dependence Task. The case against MD1 is quite simple: two entities can be modally correlated without one being dependent on the other (Fine 1995). Thus, the fact that X is modally correlated with Y does not establish that X depends on Y or vice versa. For example, on the assumption that numbers exist necessarily, marriage exists only if the number two exists. But it is not the case that the existence of marriage depends on the number two. Similarly, some particular dollar bill, d, exists only if the set having d as its sole member does. But the existence of d does not depend on the set which contains d alone—to the contrary, the existence of the set depends on the existence of the dollar bill (Fine 1995).
According to MD1, social kinds are modally correlated with collective acceptance: in every possible world in which something is K, we collectively accept that some conditions, c1…cn, are sufficient for being K. However, fact that K is modally correlated with our mental states in this way does not establish that K depends on them. Thus, MD1 does not accomplish the Dependence Task.
Indeed, these considerations demonstrate that no modal existential analysis of mind-dependence can accomplish the Dependence Task. Tinkering with the details of the proposed modal-existential relation does nothing to address this general problem. For example, the nature and content of the relevant mental states does not make a difference. That is, it does not matter whether the relevant mental states are acceptances, beliefs or something else. It does not matter whether the relevant mental states are collective or individualistic. It does not matter whether the relevant mental states are about the kinds, their instances, both or neither. Although modal correlation is (defeasible) evidence of the presence of a dependence relation, it does not constitute such a relation. It is possible that something is F only if Y exists, but being F does not depend on Y. It is precisely this point which motivates various non-modal analyses of ontological dependence (Fine 1995, 1994; Koslicki 2012, 2013; Tahko and Lowe 2016), including much of the recent work on metaphysical grounding.13
Given this, one possibility is that the collective acceptance view can be revised by appeal to the notion of ground (Griffith 2018a, 2018b; Epstein 2013, 2014, 2015a, 2015b, 2016; Schaffer 2017).14 Grounding is a non-causal relation of dependence that holds between more and less fundamental entities. Typical examples of grounding include the relationship that holds between a conjunction and its conjuncts, and the relationship that holds between a truth and its truthmakers.15
The precise nature of grounding as well as the logical form of grounding statements are subject to dispute (indeed, contrary to the way in which grounding is introduced above, one might maintain that grounding claims ought to be expressed using a sentential operator rather than a relational predicate).16 For present purposes, however, these details are immaterial. For ease of exposition, I will treat grounding as a relational predicate, and I will take the relata of the grounding relation to be facts:
(MD1-G) A kind, K, is mind-dependent =df The fact that an entity, x, is K is grounded in the fact that we collectively accept that some conditions, c1…cn, suffice for being K.17
I will argue that, even if MD1-G accomplishes the Dependence Task, it does not accomplish the Anti-Realist Task. In general, the fact that some facts (social or otherwise) are grounded in facts about our mental states (collective or otherwise) does not establish that the former are not real. Consider: the fact that I know that Sacramento is the capital of California is (partially) grounded in the fact that I believe that Sacramento is the capital of California. However, it does not follow that my belief that Sacramento is the capital of California is not real.18 Thus, revising the collective acceptance view by appeal to the notion of ground does not support social kind anti-realism.
Summing up, in this section I have argued that MD1 does not accomplish the Dependence Task, and that MD1-G does not accomplish the Anti-Realist Task. As such, neither MD1 nor MD1-G supports social kind anti-realism.
Although Thomasson explicitly endorses MD1, her motivation for endorsing the collective acceptance view suggests an alternative, essentialist interpretation of the relation that is supposed to obtain between social kinds and our mental states. Given that MD1 does not accomplish the Dependence Task and given that MD1-G does not accomplish the Anti-Realist Task, the essentialist interpretation of mind-dependence implicit in Thomasson’s argument for the collective acceptance view is worth exploring. In this section, I argue that this essentialist relation does not accomplish the Dependence Task either and therefore does not support social kind anti-realism.
According to Thomasson, ‘the minimal core of ontological realism is the position that something exists independently of all mental states’ (2003b: 581). However, she maintains that ‘a more robust realism is…often thought to require that there be a world that not only exists, but also has a certain structure independently of the mental’ (2003b: 582). On this more ‘robust’ understanding of realism, a kind, K, is real when it has its boundaries ‘independently of how our concepts and representations might happen to divide things up, in particular, independently of what we believe about the conditions relevant to drawing those boundaries’ (2003b: 582).
Thomasson argues that the way in which social kinds like money depend on our mental states means that their boundaries are not independent of our concepts and representations in this way. On her view, social kinds do not have their boundaries independently of our beliefs about where those boundaries are located because our mental states ‘play a stipulative role in constituting’ their nature (2003b: 590). Thus, although she claims to endorse a modal-existential analysis of mind-dependence, these remarks suggest an essentialist construal of the relevant dependence relation:
(MD2) A kind, K, is mind-dependent =df We determine which properties are essential to being K.19
For example, according to MD2, the social kind money is mind-dependent because we determine which properties are essential to being money. Indeed, the MD2 comports well with the remarks by both Hayek—social kinds ‘are what people think they are’ (Hayek 1943: 3)—and Searle—social kinds ‘only are what they are, because that is what we believe that they are’ (Searle 2007: 4), cited in the introduction of this paper. Both of these remarks have an essentialist flavor.
The essence of X is what X is, or what it is to be X. Essences come in two varieties: individual and general. The essential properties of an individual are what it is to be that very individual, as opposed to some other individual. For example, the essence of Queen Elizabeth II (say, her biological origin) specifies what it is to be Queen Elizabeth II, as opposed to Queen Elizabeth I, or Queen Victoria. The essential properties of a kind, K, specify what it is to be that kind, as opposed to some other kind, K*. For example, the general essence of water specifies what it is to be water, as opposed to gold or tigers. According to this interpretation of MD2, a social kind, K, is mind-dependent when we determine the general essence of being K, i.e., what it is to be K.
How do we determine the essential properties of social kinds? Thomasson indicates that we do so by the way in which we use our words and apply our concepts (2003b). For example, we determine which properties are essential to money by the way in which we use the word ‘money’ and apply the concept MONEY.20
Understood in this way, MD2 promises to substantiate the anti-realist’s idea that social kinds are merely a ‘projection’ of our thoughts onto mind-independent reality. According to MD2, social kinds do not exist ‘out there’ independently of what we may think about them. Rather, we determine which social kinds there are by imposing our concepts onto a world that is otherwise indifferent to our conceptual scheme.
However, MD2 does not accomplish the dependence task: we cannot determine the essential properties of any kind, whether or not it is social. Kinds are individuated by their essential properties.21 Suppose that a kind, K, is essentially F. If K is essentially F, then any kind that is not F is not K. It follows that K is necessarily F. If K is necessarily F, then we have no control over whether K is F, for we have no control over which properties K has necessarily. The same goes for any property that is essential to being K. If we have no control over which properties are essential to being K (because we have no control over which properties K has necessarily), then we cannot determine which properties are essential to being K— a fortiori we cannot determine which properties are essential to being K by the way in which we use our terms and apply our concepts. Thus, MD2 does not identify the way in which social kinds depend on our mental states. In other words, MD2 does not accomplish the Dependence Task. As such, it fails to support social kind anti-realism.
It might be argued that this line of argument in the preceding section is too quick. In this section, I want to reconsider MD2 by appeal to a framework devised by Iris Einheuser to defend what she calls ‘conventionalism,’ the view that ‘some aspects of reality depend on our conventions’ (2006: 460). A conventionalist about C says that ‘the way things stand C-wise’ is conventional, and that if our conventions had differed, then ‘the way things stand C-wise’ would have differed as well (2006: 459).
MD2 expresses a conventionalist view of social kinds. According to MD2, the essential properties of social kinds depend on our conventions—in particular, they depend on our linguistic or conceptual conventions. If our linguistic or conceptual conventions had been different, then the essential properties of social kinds would have been different as well.22
Although I agree that social kinds depend on our conventions in some way, I do not think that MD2 correctly identifies the sense in which social kinds are conventional. Certainly, it is conventional whether any social kinds are instantiated. For example, the fact that some people are permanent residents of the United States depends on there being various social conventions. Many millennia ago, there were no permanent residents because the relevant social conventions were not in place. Moreover, social conventions may change in the future such that there ceases to be any permanent residents.
Similarly, it is matter of convention that bills issued by the Bureau of Engraving and Printing are money—indeed, it is a matter of convention that anything (e.g., bits of paper, gold coins, cowry shells, etc.) is money. Many millennia ago, there were different social conventions, and different kinds of objects were money. Moreover, social conventions may change such that bills issued by the Bureau of Engraving and Printing cease to be money. Thus, whether anything is money depends on there being certain social conventions. Moreover, which objects are money is also a matter of social convention.
However, MD2 says something much stronger than this. According to MD2, it is not merely that the existence or instantiation of social kinds is conventional. That is, MD2 does not merely say that whether there are any women, or men, or money depends on there being certain social conventions. Rather, MD2 says that the very nature of these kinds is conventional. For example, MD2 says that it is conventional that money is a commonly used medium of exchange and measure and store of value. I will argue that social kinds are not conventional in this sense.
My argument in this section will parallel the argument given in the previous one: for any kind, K, if being F is essential to being K, then, K is necessarily F. If K is necessarily F, then it is not in our power to determine whether K is F—a fortiori it is not in our power to determine whether K is F by the way in which we use our words and concepts. Therefore, the essential properties of social kinds are not determined by linguistic or conceptual conventions.
To defend conventionalism, Einheuser calls attention to two ways in which modal claims can be evaluated. First, when evaluating modal claims we sometimes hold our linguistic and conceptual conventions fixed and consider possible worlds that differ from the actual world with respect to which properties are instantiated. For instance, consider the claim that water is necessarily H2O. When evaluating this modal claim, sometimes we envision a possible world, w, in which the clear, potable liquid that fills the lakes and rivers, and falls from the sky, has the chemical composition XYZ, rather than H2O. Although inhabitants of w might call XYZ by a word which is homonymous with our word ‘water,’ XYZ is not water, and our concept WATER doesn’t apply to it (Putnam 1973, 1975; Burge 1986). A kind, K, is water only if K is H2O.23 In other words, water is necessarily H2O.
Second, instead of varying the properties instantiated at some possible world, w, we can hold those properties fixed, and vary the way in which words, and the concepts they express, are applied to the properties instantiated at w (Burge 1979, Einheuser 2006). Following Einheuser, call the raw material the world provides—that is, the manifold of relevantly mind-independent properties—the substratum (Einheuser 2006: 461). The substratum, s, of a world, w, includes those properties which do not depend (in the relevant sense) on the way in which we apply our concepts.
According to Einheuser, a world consists not just of a substratum, but of a substratum plus a carving. A carving, c, is an abstraction from the actual conceptual conventions of the world in question. Thus, each world, w, is represented by substratum-carving pair <s,c>. The actual world, w@, is represented by <s@,c@> (Einheuser 2006: 462–463). Einheuser argues that modal claims are evaluated along both of these dimensions. Given this two-dimensional framework, MD2 can be understood as follows.
There is some world, w′ = <s@, c′>, such that the properties which are essential to money in the actual world, w@, are not essential to money in w′. In the actual world, when c@ is applied to s@, the properties of being a commonly used medium of exchange is essential to money. By contrast, in world w′, when c′ is applied to s@, the properties of being a commonly-used medium of exchange is not essential to money. Rather, some other property, say, the property of being yellow, is essential to being money. Thus, and in accordance with MD2, which properties are essential to money co-vary with our conceptual practices.
I will argue that Einheuser’s two-dimensional framework does not support MD2. To see why, it is important not to be misled by our use of the word ‘money’ in talking about the kind, K1, denoted by that word-form in the actual world, w@, and the kind, K2, denoted by the homonymous word-form ‘money’ in the possible world, w′. Although we use the same-shaped and same-sounding symbol—viz., ‘money’—in talking about both K1 and K2, we are actually talking about two different kinds—not the same kind but with different essential properties.
The properties of being a commonly used medium of exchange and being a measure and store of value are essential to kind, K1. However, by hypothesis, these properties are not essential to kind K2. Likewise, suppose that the property of being yellow is essential to kind K2 but not K1. Consequently, K1 ≠ K2. Thus, in worlds w@ and w′, speakers’ use of the word-form ‘money’ expresses different concepts and refers to different kinds. Speakers do not refer to the same kind, with different essential properties in different worlds. That is, world w′ = <s@, c′> is not a world in which kind, K1, (i.e., money) has different essential properties, as determined by our conceptual conventions in that world. Rather, w′ is a world in which speakers refer to a different kind (i.e., K2) by a word which is homonymous with our word-form ‘money,’ and express a different concept by their use of it.
Therefore, thesis that we determine which properties are essential to social kinds like money by the way in which we use social kind terms or concepts—even when regimented by Einheuser’s two-dimensional framework—is false. The essential properties of money do not vary in accordance with our conceptual practices. In the actual world, w@, we have the concept MONEY1, which refers to K1. In the possible world, w′, we have the concept MONEY2, which refers to K2. Thus, K1 and K2 are different kinds that have different essential properties, not the same kind with different essential properties.
I think that those who are inclined to endorse MD2 mistakenly conflate the metasemantic thesis that the kinds to which our words refer is determined by our linguistic or conceptual conventions (i.e., which kinds our words refer to or which concepts our terms express), with the metaphysical thesis that the essential properties of the kinds themselves are so determined. The metasemantic thesis is true but uninteresting, and the metaphysical claim is false (for the reasons outlined above). Thus, MD2 does not identify the way in which social kinds depend on our mental states. In other words, MD2 does not accomplish the Dependence Task. In that case, MD2 does not support social kind anti-realism.
In the preceding sections, I considered two prominent ways of analyzing the relation that is supposed to obtain between social kinds and our mental states. I argued that each proposal fails to accomplish the Dependence Task. In this section, I consider an alternative relation of mind-dependence drawn from Ásta’s work on the metaphysics of social properties (2008, 2013, 2018).
Consider the social kind permanent resident. What it is to be a permanent resident is to be authorized to live and work in a country indefinitely, without being a citizen of that country. Individuals ostensibly acquire the essential properties of being a permanent resident by agreement or decree. Specifically, someone is a permanent resident of say, the United States, because the appropriate individual(s) at the United States Citizenship and Immigration Services (USCIS) agree or declare that she is authorized to live and work in the United States permanently, although she is not a U.S. citizen.
Following Ásta, we may call properties such as the property of being a permanent resident conferred properties. A conferred property is one that is instantiated in virtue of subjects’ attitudes toward the entities that instantiate them. Ásta takes her cue from Euthyphro: if an action is pious because it is loved by the gods, then being pious is a conferred property (Ásta 2008). Kinds defined in terms of conferred properties (i.e., conferred property kinds) are mind-dependent in the following sense:
(MD3) A kind, K, is mind-dependent =df Being F is a conferred property and being F is essential to being K.
According to MD3, a kind, K, is mind-dependent when at least one of its essential properties is a conferred property. In other words, according to MD3, K is mind-dependent when something is K because it instantiates a property, F, and F is both (i) essential to being K, and (ii) a conferred property.
I have already suggested one social kind that is plausibly defined in terms of a conferred property, namely, being a permanent resident. Astá defends the thesis that the property of being a man and the property of being a woman are conferred properties (Ásta 2013, 2018). On her view, an individual is a woman in a context, C, because we perceive her to be a woman in that context.24 Specifically, the property of being a woman is conferred on a subject, S, if S is perceived to have some other property, which Ásta calls the base property (in her earlier work, these properties were called “grounding properties”).
For example, consider a context in which gender identity serves as the socially significant base property, e.g., in trans-inclusive subcultures. In such contexts, whether the property of being a woman is conferred on someone depends on whether others perceive them as identifying as a woman. To understand how the property of being a woman is conferred in this context, Ásta provides the following schema:
Conferred property: being a woman.
Who: those with standing in a trans-inclusive subculture, C.
What: the perception that a subject, S, has the property of identifying as a woman.
When: in a trans-inclusive subculture, C.
Base property: the property of identifying as a woman.
On Ásta’s view, if those with standing (e.g., social influence or power) in C perceive that S has the property of identifying as a woman, then the property of being a woman is conferred on S in C.
If some social kinds are conferred property kinds, then MD3 accomplishes the Dependence Task. Does it accomplish the Anti-Realist Task as well? I think not. Though it is tempting to regard conferred properties, and conferred property kinds, as metaphysically second rate, this is not yet a reason to believe that they are unreal. Next, I consider two ways of substantiating the intuition that conferred property kinds are not real. I argue that neither proposal is successful. Thus, even if MD3 accomplishes the Dependence Task, it does not accomplish the Anti-Realist Task. As such, it does not support social kind anti-realism.
The first way of spelling out the intuition of unreality concerning conferred properties is by appeal to Searle’s distinction between ontologically objective and ontologically subjective features of the world (Searle 1995). Whereas ontologically objective features are supposed to be features of the world ‘as it is in itself,’ ontologically subjective features are features as the world ‘as it is for us.’ On this view, conferred properties and conferred property kinds are ontologically subjective. They are merely features of the world ‘as it is for us.’
Although it is not entirely clear how Searle intends to draw the distinction between features of the world that are ontologically subjective and those that are ontologically objective, he is very clear that mental states like pain are ontologically subjective (Searle 1995: 8). But it does not follow from the fact that pain is ontologically subjective that pain is not real. If some creature is in pain, its being in pain is no less real than its being human, its being bipedal, or its weighing 120lbs. Indeed, if mental states are type or token identical to physical states of the brain, then there is even less reason to regard them as unreal. So, even granting the assumption that conferred properties are ontologically subjective, this does not demonstrate that they are not real.25
The second way of capturing the idea that conferred property kinds are not real is by appeal to the idea that conferred properties are response-dependent (Ásta 2008, Passinsky forthcoming). A response-dependent property, F, is a particular kind of dispositional property: F is the property of being disposed to produce a certain psychological response (cognitive or affective) in the relevant creatures under certain conditions (i.e., the conditions under which the dispositional property is manifested).26
For example, to say that the property of being red is response-dependent is to say that redness is the property of being disposed to produce reddish sensations in us under certain conditions. Likewise, to say that the property of being a woman is response-dependent is to say that being a woman is the property of being disposed (under the relevant conditions) to produce the perception (in the relevant subjects) that some individual is a woman.
The idea that response-dependent properties like being red do not have the same metaphysical status as response-independent properties like being hexagonal is a tempting one. If conferred properties are response-dependent properties, and response-dependent properties are less than fully real, then this is a reason to believe that conferred property kinds are less than fully real as well.
However, the fact that some properties are identical to dispositions to produce a psychological response does not obviously diminish their reality in any way (Rosen 1994, Egan 2006). To the contrary, such properties seem to be metaphysically on par with dispositional properties which do not implicate any mental states, such as the property of being disposed to shatter. This is because the relevant psychological responses, as well as the objects that are disposed to produce them (along with the relevant manifestation conditions), are plausibly physical entities, and there is no reason to believe that physical entities (e.g., brain states and human organisms) are unreal.
Thus, even if conferred-properties are response-dependent, this does not provide us with any reason to doubt the reality of conferred properties or the corresponding kinds. In other words, even if MD3 accomplishes the Dependence Task, it does not accomplish the Anti-Realist Task. As such, it does not support social kind anti-realism.
I have argued against the view that social kinds are not real in virtue of being mind-dependent. I surveyed several ways of defining the relation that is supposed to obtain between social kinds and our mental states and argued that each proposal either fails to accomplish the Dependence Task or fails to accomplish the Anti-Realist Task. As such, social kind anti-realism is not well justified: anyone who wishes to defend social kind anti-realism must provide an alternative explanation of how social kinds depend on our mental states in a way that impugns their reality. In the absence of such an explanation, there is no reason to endorse social kind anti-realism. To the contrary, the foregoing arguments should lead us to be suspicious of the idea that social kinds are unreal because they depend on our mental states.27
1An anonymous referee suggests that there may be different kinds of social kinds, and that they may depend on our mental states in a diversity of ways. For example, Amie Thomasson (2003b) distinguishes between social kinds that are “epistemically opaque” (e.g., racism, recessions) and those that are not (e.g., money). Likewise, Khalidi (2015) distinguishes between three kinds of social kinds. My arguments are directed at the more general claim that social kinds simpliciter are unreal for two reasons. First, Searle’s defense of social kind anti-realism includes all social kinds, and not just a subset of them. Second, the question of whether there are distinctions between different kinds of social kinds is orthogonal to the question of whether social kind anti-realism is well founded. For the purposes of the argument in this paper, it does not matter whether there are different kinds of social kinds, or whether different social kinds depend on our mental states in different ways. I argue that there is no reason to think that any social kinds are unreal in virtue of being mind-dependent.
4Social kind anti-realism is a prominent view in the literature, but it is not the only view. Several of the essays collected in Haslanger (2012), as well as Guala (2014), Khalidi (2010, 2013, 2015, 2016); Mallon (2003, 2016); and Root (2000) defend the thesis that social kinds are mind-dependent, but not that they are unreal.
5Khalidi (2016) considers several varieties of mind-dependence that are importantly different from those I consider in what follows. According to Khalidi, whether or not something is mind-dependent is irrelevant to the question of whether that thing is real (225). I am very sympathetic with this claim. The arguments in this paper provide further reasons to support Khalidi’s conclusion.
6One option is to define realism and anti-realism are defined in terms of mind-independence and mind-dependence respectively. However, if ‘real’ and ‘unreal’ are simply synonyms for ‘mind-independent’ and ‘mind-dependent,’ social kind anti-realism fails to be an interesting thesis. It is uncontroversial that social kinds are mind-dependent and therefore unreal according to this definition.
7Though I argue that social kind anti-realism is not well-justified, we should not therefore conclude that social kinds are real after all. In other words, my argument against social kind anti-realism is not intended to support social kind realism.
8According to Amie Thomasson, the claim that some kind, K, is not real is not equivalent to the claim that K does not exist. Thomasson says this quite explicitly: ‘On this understanding, denying a relative ontological realist thesis regarding things of kind K should not automatically be equated with denying that there are entities of kind K, or else we could not distinguish, e.g., conceptualism or constructivism from eliminativism’ (2003b, 582 fn. 5). The arguments presented in what follows do not concern those philosophers who argue that races do not exist (see Appiah 1995, 1996 and Zack 1993, 2002). Rather, the main targets of my critique are those philosophers who maintain that a social kind, K, exists but is not real. Thanks to Kate Ritchie on this point.
9One species of mind-dependence that I will not consider is causal mind-dependence. This is because proponents of social kind anti-realism do not argue that social kinds fail to be real because our thoughts cause them to exist. Indeed, according to Searle, mere causal mind-dependence does not justify the claim that social kinds are not real (Searle 1995, 156). Instead, proponents of social kind anti-realism argue that the relation that obtains between social kinds and our mental states (the Dependence Task), and which undermines their reality (the Anti-Realist Task), is a non-causal relation of some kind. Thus, in what follows I consider only non-causal relations of mind-dependence. Others also deny that causal mind-dependence entails anti-realism (see Egan 2006, Haslanger 1995, Khalidi 2016, and Rosen 1994). I follow Sally Haslanger and Ron Mallon in thinking that discursively constructed kinds (i.e., kinds subject to what Ian Hacking (1996) calls ‘looping effects’) are causally mind-dependent in this sense (see Cooper 2004, Khalidi 2010, 2013, 2016; and Mallon 2003, 2016 for germane discussions of discursively constructed or ‘interactive’ kinds).
10According to Thomasson (2003a) and Khalidi (2015), only some social kinds depend on our mental states in this way. They argue that many social kinds (e.g., recessions, racism) do not depend on our having mental states that are about the kinds in question. For example, it is not the case that racism and recessions exist only if we have thoughts about racism and recessions. Rather, these kinds are mind-dependent in the sense that they exist only if certain other mental states exist. This relation, unlike the one under consideration in this section, does not specify the nature and content of the mental states on which the relevant kinds depend. However, because it is a modal-existential relation, the arguments presented in this section undermine it as well.
11Searle’s original formulation of what I am calling MD1 is that a social kind, K, exists only if we collectively accept a constitutive rule with the following logical form: X (some conditions) counts as Y (the kind in question, or what Searle would call the ‘status function’) in C (the relevant context). I think that it is plausible to interpret Searle’s constitutive rules as supplying sufficient conditions for the existence of instances of social kinds (i.e., some individual being a member of the kind in question), and so I think that MD1 is faithful to Searle’s view. Thomasson (2003b) likewise interprets Searle in this way.
12Francesco Guala (2010, 2014) also argues that collective acceptance of some conditions c1…cn is neither necessary nor sufficient for the existence of social kinds. However, his argument for this conclusion differs from the one I offer here in important ways. Guala argues that MD1 is false, whereas I argue that MD1 is not a genuine dependence relation. According to Guala, the conditions c1…cn that we collectively accept as sufficient for being money are coordination devices which facilitate the requisite cooperative behavior which is necessary and sufficient for the existence of money.
15Work by Audi (2012), Fine (2012), Koslicki (2013), Raven (2012), Rosen (2010), Schaffer (2009), and Wilson (2014) provide some early discussions of grounding and inspired debate concerning the existence and nature of ground.
17Those who reject the idea that grounding holds between facts, or those who believe that grounding claims ought to be expressed using a sentential operator, are invited to reformulate MD1-G accordingly.
18NB: Kit Fine (2001) argues that grounded entities are not real. However, this claim is not equivalent to social kind anti-realism. First, as I argue in section 2.2, social kind anti-realists typically do not (and need not) accept the claim that grounded entities are unreal. To the contrary, social kind anti-realists typically accept that many nonfundamental kinds (e.g., atoms, potassium, tigers) are real. Indeed, proponents of social kind anti-realism typically argue that physical, chemical, biological, and even psychological kinds are real, but many social kind are not. This is what Francesco Guala calls the ‘difference thesis’ (2014). Second, in the idiom of grounding, social kind anti-realism is the thesis that social kinds are not real because they are grounded in our mental states. By contrast, Fine’s claim, applied to social kinds, is that social kinds are not real because they are grounded, i.e., non-fundamental. In other words, according to social kind anti-realism, the reason why social kinds are unreal is not simply that they are dependent; rather, it is that they are mind dependent. Indeed, the truth of Fine’s claim is compatible with the falsity of social kind anti-realism. In that case, social kind anti-realists may have reached the right conclusion for the wrong reasons—their argument is invalid. This is what I hope to show in this paper: being mind-dependent does not entail being unreal. But social kinds may be unreal (in some sense) for different reasons.
19A similar claim is also defended by Ásta (2008b), who argues that we determine which properties are essential to an object. Thanks to an anonymous referee for making this connection to Ásta’s anti-realist account of essentiality.
21An anonymous reviewer worries that I am not entitled to assume that social kinds are individuated by their essential properties. However, kinds are individuated by the properties that identify them. Every kind that exists has an identity (i.e., a nature), and the essential properties of a kind are just those properties that identify the kind in question. If one does not wish to call the properties that identify kinds ‘essential properties,’ that’s fine. However, this is not a substantive metaphysical disagreement; it is a terminological one. My argument would go through just the same if we were to call the properties that identify kinds ‘schmessential properties’ or ‘identifying properties.’ Thus, I do not believe that my claim that kinds are individuated by their essential properties is problematic.
22Ásta (2008b) defends a view of essentiality that is similar to conventionalism insofar as it aims to deliver the result that we determine which properties are essential to an object by how we use our concepts. According to Ásta, ‘the property of being an essential property of an object—essentiality—is conferred. I maintain that essentiality is conferred by our use of concepts’ (138). However, Ásta also notes that her view differs from conventionalism in important ways (144–147). Thanks to an anonymous referee for identifying this connection between Ásta’s and Einheuser’s frameworks.
23Admittedly, this conclusion is controversial. However, it is merely meant to illustrate the point that normally, when evaluating metaphysical possibility, we make our judgments on the basis of our actual conceptual practices. On the assumption that the way in which we actually apply our concepts is such that we judge that a liquid is water if and only if it is H2O, water is necessarily H2O.
24Charles Mills defends a similar view with respect to race. Mills argues that individuals acquire the property of being black or being white in virtue of the fact that we intersubjectively judge them to be black or white (Mills 1998).
25Moreover, if, as I suspect, Searle’s distinction between ontologically subjective and ontologically objective features of the world simply corresponds to the distinction between mind-dependent and mind-independent features, then the claim that conferred property kinds are ontologically subjective is equivalent to the claim that those kinds are mind-dependent and does not offer any independent reason for thinking that such kinds are unreal.
26Many discussions of response-dependence concern response-dependent concepts, see Johnston (1989), Pettit (1991), Wedgwood (1997), and Wright (1992). For brevity and clarity, however, I focus here on response-dependent properties instead.
The author has no competing interests to declare.
Ásta. 2008a. Siding with Euthyphro: Response-Dependence and Conferred Properties. European Journal of Philosophy, 18(1): 108–125. DOI: https://doi.org/10.1111/j.1468-0378.2008.00303.x
Ásta. 2008b. Essentiality Conferred. Philosophical Studies, 140: 135–148. DOI: https://doi.org/10.1007/s11098-008-9230-4
Ásta. 2013. The Social Construction of Human Kinds. Hypatia, 28(4): 716–732. DOI: https://doi.org/10.1111/j.1527-2001.2012.01317.x
Ásta. 2018. The Categories We Live By. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/oso/9780190256791.001.0001
Audi, P. 2012. Grounding: Toward a Theory of the “In-Virtue-of” Relation. The Journal of Philosophy, 109(12): 685–711. DOI: https://doi.org/10.5840/jphil20121091232
Burge, T. 1979. Individualism and the Mental. Midwest Studies in Philosophy, 4(1): 73–122. DOI: https://doi.org/10.1111/j.1475-4975.1979.tb00374.x
Burge, T. 1986. Intellectual Norms and Foundations of Mind. The Journal of Philosophy, 83(12): 697–720. DOI: https://doi.org/10.2307/2026694
Cooper, R. 2004. Why Hacking is Wrong about Human Kinds. British Journal for the Philosophy of Science, 55(1): 73–85. DOI: https://doi.org/10.1093/bjps/55.1.73
Egan, A. 2006. Secondary Qualities and Self-Location. Philosophy and Phenomenological Research, 72(1): 97–119. DOI: https://doi.org/10.1111/j.1933-1592.2006.tb00492.x
Einheuser, I. 2006. Counterconventional Conditionals. Philosophical Studies, 127: 459–482. DOI: https://doi.org/10.1007/s11098-004-7790-5
Epstein, B. 2013. Social Objects without Intentions. In: Konzelmann Ziv, A and Schmid, HB (eds.), Institutions, Emotions, and Group Agents: Contributions to Social Ontology. Dordrecht: Springer. pp. 53–68. DOI: https://doi.org/10.1007/978-94-007-6934-2_4
Epstein, B. 2014. What Is Individualism in Social Ontology? Ontological Individualism vs. Anchor Individualism. In: Zahle, J and Collin, F (eds.), Rethinking the Individualism-Holism Debate. Dordrecht: Springer. pp. 17–38. DOI: https://doi.org/10.1007/978-3-319-05344-8_2
Epstein, B. 2015a. How Many Kinds of Glue Hold the Social World Together? In: Gallotti, M and Michael, J (eds.), Perspectives on Social Ontology and Social Cognition. Dordecht: Springer. pp. 41–55. DOI: https://doi.org/10.1007/978-94-017-9147-2_4
Epstein, B. 2015b. The Ant Trap: Rebuilding the Foundations of the Social Sciences. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/acprof:oso/9780199381104.001.0001
Epstein, B. 2016. A Framework for Social Ontology. Philosophy of the Social Sciences, 46(2): 147–167. DOI: https://doi.org/10.1177/0048393115613494
Fine, K. 1994. Essence and Modality. Philosophical Perspectives, 8: 1–16. DOI: https://doi.org/10.2307/2214160
Fine, K. 1995. Ontological Dependence. Proceedings of the Aristotelian Society, 95: 269–290. DOI: https://doi.org/10.1093/aristotelian/95.1.269
Fine, K. 2001. The Question of Realism. Philosophers’ Imprint, 1(1): 1–30. DOI: https://doi.org/10.1093/aristotelian/95.1.269
Fine, K. 2012. Guide to Ground. In: Correia, F and Schnieder, B (eds.), Metaphysical Grounding Cambridge: Cambridge University Press. pp. 37–80. DOI: https://doi.org/10.1017/CBO9781139149136.002
Guala, F. 2010. Infallibilism and Human Kinds. Philosophy of the Social Sciences, 40(2): 244–264. DOI: https://doi.org/10.1177/0048393109343114
Guala, F. 2014. On the Nature of Social Kinds. In: Gallott, M and Michael, J (eds.), Perspectives on Social Ontology and Social Cognition. Dordrecht: Springer. pp. 57–68. DOI: https://doi.org/10.1007/978-94-017-9147-2_5
Griffith, AM. 2018a. Social Construction and Grounding. Philosophy and Phenomenological Research, 97(2): 393–409. DOI: https://doi.org/10.1111/phpr.12376
Griffith, AM. 2018b. Social Construction: Big-G Grounding, Small-g Realization. Philosophical Studies, 175(1): 241–260. DOI: https://doi.org/10.1007/s11098-017-0865-x
Hacking, I. 1996. The Looping Effects of Human Kinds. In: Sperber, D, Premack, D and Premack, AJ (eds.), Causal Cognition: A Multidisciplinary Debate. Oxford: Oxford University Press. pp. 351–383. DOI: https://doi.org/10.1093/acprof:oso/9780198524021.003.0012
Haslanger, S. 1995. Ontology and Social Construction. Philosophical Perspectives, 23(2): 95–125. DOI: https://doi.org/10.5840/philtopics19952324
Haslanger, S. 2012. Resisting Reality: Social Construction and Social Critique. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/acprof:oso/9780199892631.001.0001
Hayek, F. 1943. The Facts of the Social Sciences. Ethics, 54(1): 1–13. DOI: https://doi.org/10.1086/290368
Johnston, M. 1989. Dispositional Theories of Value. Proceedings of the Aristotelian Society, 63(1): 139–174. DOI: https://doi.org/10.1093/aristoteliansupp/63.1.89
Khalidi, MA. 2010. Interactive Kinds. British Journal for the Philosophy of Science, 61(2): 335–360. DOI: https://doi.org/10.1093/bjps/axp042
Khalidi, MA. 2013. Natural Categories and Human Kinds. Cambridge: Cambridge University Press. DOI: https://doi.org/10.1017/CBO9780511998553
Khalidi, MA. 2015. Three Kinds of Social Kinds. Philosophy and Phenomenological Research, 90(1): 96–112. DOI: https://doi.org/10.1111/phpr.12020
Khalidi, MA. 2016. Mind-Dependent Kinds. Journal of Social Ontology, 2(2): 223–246. DOI: https://doi.org/10.1515/jso-2015-0045
Koslicki, K. 2012. Varieties of Ontological Dependence. In: Correia, F and Schnieder, B (eds.), Metaphysical Grounding: Understanding the Structure of Reality. Cambridge: Cambridge University Press: pp. 186–305. DOI: https://doi.org/10.1017/CBO9781139149136.008
Koslicki, K. 2013. Ontological Dependence: An Opinionated Survey. In: Schnieder, B, Hoeltje, M and Steinberg, A (eds.), Varieties of Dependence: Ontological Dependence, Grounding, Supervenience, Response-Dependence. Munich: Philosophia Verlag. pp. 21–64.
Mallon, R. 2003. Social Construction, Social Roles, and Stability. In: Schmitt, FF (ed.), Socializing Metaphysics: The Nature of Social Reality. Lanham: Roman & Littlefield Publishing, Inc. pp. 327–353.
Mallon, R. 2016. The Construction of Human Kinds. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/acprof:oso/9780198755678.001.0001
Pettit, P. 1991. Realism and Response-Dependence. Mind, 100(4): 587–626. DOI: https://doi.org/10.1093/0199251878.003.0002
Putnam, H. 1973. Meaning and Reference. The Journal of Philosophy, 70(19): 699–711. DOI: https://doi.org/10.2307/2025079
Raven, MJ. 2012. In Defence of Ground. Australasian Journal of Philosophy, 90(4): 687–701. DOI: https://doi.org/10.1080/00048402.2011.616900
Raven, MJ. 2015. Ground. Philosophy Compass, 10(5): 322–333. DOI: https://doi.org/10.1111/phc3.12220
Root, M. 2000. How We Divide the World. Philosophy of Science, 67(3): 628–639. DOI: https://doi.org/10.1086/392851
Rosen, G. 1994. Objectivity and Modern Idealism: What is the question? In: O’Leary-Hawthorne, J and Michael, M (eds.), Philosophy in Mind. Dordrecht: Klewer. pp. 277–319. DOI: https://doi.org/10.1007/978-94-011-1008-2_17
Rosen, G. 2010. Metaphysical Dependence: Grounding and Reduction. In: Hale, B and Hoffmann, A (eds.), Modality: Metaphysics, Logic, and Epistemology. Oxford University Press. pp. 109–136. DOI: https://doi.org/10.1093/acprof:oso/9780199565818.003.0007
Schaffer, J. 2017. Social Construction as Grounding. Philosophical Studies, 174(10): 2449–2465. DOI: https://doi.org/10.1007/s11098-016-0738-8
Searle, J. 2008. Language and Social Ontology. Theory and Society, 37(5): 443–459. DOI: https://doi.org/10.1007/s11186-008-9068-y
Searle, J. 2010. Making the Social World: The Structure of Human Civilization. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/acprof:osobl/9780195396171.001.0001
Searle, J. 2014. Are there social objects? In: Gallotti, M and Michael, J (eds.), Perspectives on Social Ontology and Social Cognition. Dordrecht: Springer. pp. 17–26. DOI: https://doi.org/10.1007/978-94-017-9147-2_2
Thomasson, A. 2003a. Foundations for a Social Ontology. Protosociology, 18/19: 269–440. DOI: https://doi.org/10.5840/protosociology200318/199
Thomasson, A. 2003b. Realism and Human Kinds. Philosophy and Phenomenological Research, 67(3): 580–609. DOI: https://doi.org/10.1111/j.1933-1592.2003.tb00309.x
Tuomela, R. 2007. The Philosophy of Sociality: The Shared Point of View. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/acprof:oso/9780195313390.001.0001
Tuomela, R. 2013. Social Ontology: Collective Intentionality and Group Agents. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/acprof:oso/9780199978267.001.0001
Wilson, JM. 2014. No Work for a Theory of Grounding. Inquiry, 57(5–6): 535–579. DOI: https://doi.org/10.1080/0020174X.2014.907542