1 Introduction

The idea that certain group entities, especially large and institutional ones such as corporations and parliaments, can possess sui generis, irreducible mental states, like beliefs, desires, intentions, and so on, has recently – and perhaps somewhat surprisingly – become popular in philosophy. The proponents of the strong realist view include List & Pettit (2011), List (2014), Pettit (2014), Tollefsen (2002, 2015), Kallestrup (2020), and Hess (2014). On this view, statements about the putative mental states of groups should be taken as describing genuine mental states of pan-individual agents. If we consider the statement “Google believes that open source is good” (as is stated by Google itself on its website), we should see this statement as describing a genuine belief that is held by this corporation. This belief is real and is not reducible to any mental states of individuals. The realist proposal has also encountered significant resistance (see, e.g., Ludwig (2017a, b; Overgaard and Salice (2021); Rupert (2005, (2014)). The opponents of realism think that is highly implausible to think that groups of individuals can genuinely possess mental states.

The debate has a long history. The issue of whether pan-individual entities can be treated as bona fide agents and bearers of mental states has been discussed in legal theory (see, e.g., Orts 2013, Ch. 2 for an overview), and in philosophy of social science, with such figures as Menger, Hayek, and Weber defending methodological individualism (see, e.g., Heath 2020 for a synopsis of the issue). One of the things that might be seen as distinguishing the current episode of this debate from previous ones is that one of the main focal issues of the current debate is a purely ontological one (namely the question of whether group minds exist); in contrast, in previous variants of the debate, the ontological question was more embedded in questions regarding the responsibility of collectives (in the case of legal theory) and the proper methodology of social sciences (including economics), as was the case in the methodological individualism debate.

Moreover, the current debate seems to centre on two issues: the first is whether group agents have all the necessary features of genuine-minded entities; the second is whether it is possible to reduce talk about the mental states of such groups to the mental states of individual subjects. In this way, the debate has focused on first-order philosophical issues, and not much interest has been given to the metaphilosophical issues pertaining to the debate (a notable exception is Himmelreich 2021).

This paper has two aims. The first is to provide reflection on the meta-metaphysical assumptions made by proponents of the most popular contemporary approaches to group mentality. The second is to provide insight into possible alternative approaches. This will serve a geographical aim by providing a tentative map of the possible options in the debate: both the currently accepted ones, and those that can serve as alternatives. Moreover, I will try to provide some tentative reasons which might motivate us to move away from the currently accepted presuppositions.

As I see it, the aim of meta-metaphysical reflection is to gain understanding of the nature of the (potentially apparent) ontological assumptions that we seem to make when engaging in a particular discourse. To give one example: meta-metaphysical reflection regarding the status of mathematics aims to help us understand what it means when we “talk about numbers”; more precisely, this meta-metaphysical reflection aims to help us understand what to make of the “ontological commitment” to numbers that we seem to make when using mathematical language. In the same vein, we might ask how to understand the implicit commitments to the existence of mentality of groups that we seem to make when we are talking about group beliefs and intentions (which we do in both folk and scientific contexts). Such a “hermeneutic” aim is definitely a modest one: there is nothing here that might have any direct impact on existing scientific practices. Meta-metaphysical reflection does not seem to directly help us answer either questions of the responsibility of groups or questions about the proper methodology of social sciences. Still, I believe that given the oft-expressed puzzlement regarding the possibility of there being genuine pan-individual minds, such reflection is warranted.

The paper proceeds as follows. First, I provide an overview of the state of the debate concerning group beliefs in contemporary analytic social ontology. Then, I turn to the issue of the meta-metaphysical assumptions of the current debate, and I argue that the two most popular approaches both share a commitment to meta-metaphysical realism. Then, I provide some motivations to look for alternatives to these meta-realist views. I develop a characterization of two such possible alternative positions that are based on rejection of meta-metaphysical realism, namely fictionalism and deflationism. The final section is devoted to a discussion of the possible criteria which can be used to choose from the theories described.

2 The state of the debate

As was noted in the previous section, the contemporary debate about institutional beliefs centres on the conflict between strong realism and eliminative reductionism. The motivation usually provided for the strong realist view is two-fold. First, it is claimed that certain popular strands in the philosophy of mind, like functionalism and intentional-stance theory, allow entities other than humans and other biological organisms to be bearers of genuine intentional states. On this line of reasoning, institutional groups can be said to be genuine believers (and subjects with intentions and the like) because our best theories about the mental deem such entities to be capable of having genuine mentality. The second motivation comes from the idea that in order to make sense of such groups’ genuine responsibility we must ascribe to them the capacity to act intentionally, and ascribing such a capacity to them requires ascribing full-fledged intentional states to them. As Tollefsen put it: “Our practice of attributing responsibility to organizations (.) seems to presuppose that organizations literally have intentional states. For we could not hold them legally and morally responsible for an act unless they intended to commit the act” (Tollefsen, 2002, p. 396).

Philosophers who oppose strong realism usually claim that we might deny that corporations, parliaments and so on have minds of their own, and these philosophers attempt to provide a reductive explanation of the attributions of beliefs to institutional groups. Perhaps the most prominent contemporary proponent of this approach is Ludwig (see Ludwig, 2017a, b), for whom attributions of beliefs to groups are shorthand descriptions of the complex reality of interactions between individuals, some of whom have special status as they act as proxy agents for other agents. This position might be called “eliminative-reductionism” as it simultaneously tries to eliminate the commitment to irreducible group minds whilst also reducing talk about them to talk about the mental states and activities of individuals.

In order to understand the dynamic of the debate, it is important to note that one of the main issues in the debate between these two camps deals with the possibility of ‘paraphrasing away’ statements that ascribe mental states to groups (this observation comes from Himmelreich (2017)). Such paraphrases should, ideally, ensure that statements about group mental states that we take to be true are indeed true, but they should also show us how this truth arises from statements about individuals. The success of the paraphrase project would give a strong reason to opt for the eliminative-reductionist proposal, whereas the impossibility of providing a satisfying paraphrase is thought to lend support to strong realism.

3 The meta-metaphysical assumptions of the debate

It might be noted that the debate, as it has been conducted so far, has focused on first-order issues. Little attention has been given to the meta-philosophical problems that might arise with relation to the question of the reality of group beliefs. This might be considered surprising, given that in many other areas of philosophy, including general metaphysics, meta issues are very much at the centre of attention.

One notable exception is Himmelreich (2021), who argues that the difference between strong realism and eliminative-reductionism is mirrored in the difference between the meta-metaphysical assumptions of these positions. On his account, realists assume a Quine-style approach, according to which we should accept into our ontology only those entities that our best theories commit us to. In this way, the hypothesis of the existence of irreducible group mental states is seen as justified by our best theories of the mental and social realms. On the other hand, the eliminativists (in Himmelreich’s interpretation) rely on the truth-maker approach to meta-metaphysics. This means that they look for more-fundamental entities to make statements about groups’ mental states true. The result is that they consider individual actions and mental states to provide these truth-makers.

This is certainly an important observation, although there might be some concerns that the division is not so neat: there are aspects of the eliminativist’s view that make it look as if it relies on Quinean meta-metaphysical assumptions (which Himmelreich also admits). My aim here, however, is not to get into a polemic with Himmelreich but to note that, in certain important respects, both the best-theory approach and the truth-maker view are similar. This is because both these views might be considered to be sub-species of a broadly realist approach to meta-metaphysics.

The most central tenet of the broadly realist approach to meta-metaphysics (as defined, for example, by Chalmers (2009)) is the idea that the job of ontology/metaphysics is to determine what really exists. Broadly conceived realism makes two basic claims about the status of the discourse about what exists: the first is that there are objectively true answers to questions about existence; the second claim is that the answers to ontological questions are non-trivial, which means that we need to engage in serious theorizing in order to arrive at the correct answers.

The debate on the reality of group mental states is conducted in a way that relies on realist assumptions. First, the debate between the realists and the eliminativists is presented as centring on the question of whether groups can have genuine beliefs, desires, and intentions. Thus, the issue is presented as concerning the objective answer to an ontological question, namely “do group beliefs and intentions exist?”. The following quote from List and Pettit might illustrate how deeply they are committed to meta-metaphysical realism:

“(…) often the ascription of agency to groups expresses a correct and important observation, both in common and in scientific discourse: a correct observation, because there really are group agents; and an important one, because to overlook their presence would be to miss out on a significant aspect of the social world” (List & Pettit, 2011, p. 4, emphasis mine).

Second, the debate heavily relies on serious theoretical questions, one of the most central of which regards the validity of certain theories of the nature of beliefs, and there is also the issue of whether mental predicates, as defined within these theories, can be adequately applied to groups. Tollefsen (2002, 2015) uses Dennett’s interpretivism (Dennett, 1987) to justify the idea that groups might really be treated as having mental states. List and Pettit (2011) rely on a certain form of functionalism to achieve the same aim. There is also the pressing question of the theoretical indispensability of attributions of mental states to groups. Both Tollefsen (2015) and List & Pettit (2011) heavily rely on the idea that without postulating genuine group mentality we miss important explanatory connections in the social realm. In this way, the debate is conceptualized as non-trivial, and the outcome is seen as depending on difficult theoretical issues.

This is also underscored by eliminativism. Many of the arguments that are made against the idea that a group can have genuine beliefs rely on the idea that groups do not have certain characteristics that, according to a well-developed theory of the mental, all genuine believers must have: for example, it is said that groups cannot have mental states as they lack genuine phenomenal consciousness, which is a prerequisite for genuine mentality (see, e.g., Overgaard & Salice 2021). Another important argument against strong realism invokes the idea that it is implausible to ascribe distinct causal powers to group minds (see Rupert 2005, 2014). However, this argument is also reliant on non-trivial philosophical assumptions, such as the claim that only causally efficacious entities and states truly exist.

Thus, both realists and eliminativists see the issue of institutional group minds as a genuinely deep, open philosophical question. They consider it possible to give an objective answer to this question if we engage in serious theorizing. It is – as the assumption goes – either objectively true that groups possess genuine intentional mental states, or it is objectively false that they do. The role of philosophical theory is to discover the facts that would make either of these answers true. In this way, both theories meet the criteria spelled out in Chalmers’s definition of meta-metaphysical realism.

4 Motivation to look for alternatives

In general philosophy the notion that we can – and indeed should – look for approaches which see the issues of ontology differently than is implied by meta-metaphysical realism has gained significant popularity in recent decades. The idea is that, by looking for an alternative understanding of the function of statements in potentially metaphysically problematic discourses, we might gain better insight into the genuine aims of the practice of ontology. Additionally, proponents of the alternative meta-metaphysical frameworks want to show that we might use such discourses without being forced to accept problematic metaphysics.

In my view, there are some significant theoretical gains to be had from broadening the set of possible stances in the debate about the reality of institutional group beliefs. First, the alternatives presented might offer ways of formulating answers to the basic question of whether group minds exist in a way that eschews at least some of the problems to which the current characterization of this debate leads. Second, looking into these alternative theories could allow us to see that answers to particular questions that were traditionally thought to go necessarily together might in fact be combined in a different way.

The main problem that stems from the way the debate is currently conceptualized is that both positions incur substantial theoretical debts in order to explain certain simple intuitions. In the case of strong realism, the intuition is that talk concerning the mental states of institutions is justified, and we cannot equate claims about these states with simple summative descriptions of individual states. However, the substantial debts that the strong realist must pay for his solution include providing the “right” metaphysics of mind and solving the exclusion problem for group mental states. Eliminative reductionism, on the other hand, starts with the simple intuition that groups do not seem to be minded entities, but this simple intuition leads to complex and potentially controversial reductive analyses. So, the natural question is whether such simple intuitions could be expressed by theories that are less theoretically burdened.

The second concern about the present state of the debate is that it lumps together answers to several in-principle-distinct questions pertaining to the status of institutional groups as believers. For example, it seems to be one issue whether such groups can truly be ascribed mentality, but whether such ascriptions might be reduced seems to be quite a different issue. One might have several reasons to be sceptical of the idea of groups having any mental states yet be unconvinced about the feasibility of individualistic reductions. Thus, it seems that the idea that only two approaches are available might be seen as somehow stifling the debate.

Searching for alternative approaches might allow us to avoid these problems. In the context of the discussion about the reality of group minds, I think that two such alternative approaches might be illuminating: fictionalism and deflationism. I do not wish to endorse either of these views but rather to present them as possible points of reference for discussion.

Obviously, moving away from the realistic approach might also be seen as having important drawbacks. The most important one is that we might lose the intuition of the “seriousness of the debate”; it is sometimes claimed (see, e.g., Chalmers 2009 for a useful overview of realists’ motivation) that only by sticking to a meta-metaphysical realist view might we save the idea that our debate regarding the ontological status of the things we are interested in is a substantial one. Rejecting this assumption might then be considered to be an attempt to avoid the possible hard questions that the issue of group mentality presents us with. An ardent supporter of meta-metaphysical realism could argue that questions of whether mental vocabulary can be legitimately extended to pan-individual entities, as well as questions of how to make sense of attributions of responsibility to groups, are important questions about non-trivial features of reality. Such a meta-realist could thus argue that adopting a non-realist meta-metaphysical view would be an attempt to avoid problems rather than solve them. Still, I believe that there might be important (and under-explored) theoretical virtues of the non-realist meta-metaphysical approaches, and this makes them worth discussing.

5 Fictionalism and group minds

The fictionalist position is not very popular in the current debate on group minds that is conducted within analytic social ontology. However, outside this sphere, fictionalist intuitions have recently been expressed, especially in legal theory and, in economics, in the theory of the firm.

In legal theory, it is customary to talk about “legal persons” as being in, a sense, “legal fictions” (see, e.g., Kurki 2019, Chap. 1), but such pronouncements by themselves do not need to signify a commitment to fictionalist theory in a philosophical sense. Still, there are important claims in legal theory which seem to assume a deeply fictionalist view on the issue of group intentionality. In the following, I will focus on one such example of the fictionalist approach to institutional mentality in legal theory, namely Canale’s (2021) theory of legislative intent.

Canale’s objective is to understand what sense can be given to an idea that is common amongst lawyers, namely that the function of legal interpretation is to discover the intention of legislature. According to Canale, it is true that, in order to enact a statute, the legislators must have some shared intentions. However, for him it is not literally true that there exists a single object, namely legislature, that is capable of having communicative intentions with definite semantic content. This is because the particular lawmakers that take part in the process of the enactment of legal statutes have distinct and often conflicting aims, motivations, and so on. Canale is also sceptical of attempts to use realistic accounts of group agency; however, for the sake of the brevity of the exposition, I will not restate his arguments here.

What matters for our purposes is his positive proposal, namely the idea that we should treat statements which attribute communicative intentions to legislatures as fictional ones. As he writes: “legal officials engage in a fictional game of make-believe when they interpret a statute according to legislative intent, although they are often not aware of that” (Canale, 2021, p. 60). Engaging in such fiction is vital for legal officials, as by fictionally assuming that there is such a thing as a single, semantically determinate, communicative intention of the legislature, the participants in this fictional game might rationally disagree over the content of law.

The idea is that when legal officials, like Supreme Court judges, engage in the process of interpreting law, it is rational for them to do so as if statutes were created by a single, rational, coherent (and morally good) group subject who is capable of having determinate intentions which are communicated by statutes. This assumption, according to fictionalism, is false, as actually existing legislatures are not groups with minds of their own that exist over and above the minds of individual legislators (who might have conflicting intentions, etc.). According to fictionalism about legislative intention, engaging in this pretense of there being a single legislative intention that determines the content of law puts important normative constraints on the process of legal interpretation. The fictional unified legislator is, in the process of legal interpretation, seen as embodying certain important values (of consistency, coherence, justice, and fairness), and these values guide the process of legal interpretation. Debates about the ‘correct’ interpretation are then just debates about how to best realize these values (Canale, 2021, esp. p. 63–64).

Canale’s proposal is important as it shows that one of the paradigmatic cases of entities that are suspected of possessing group mental states (namely parliaments) might be reasonably treated in the fictionalist manner. The arguments that Canale provides for the fictionalist approach are partly of a metaphysical and partly of a methodological nature: he is interested in the role that the ascription of genuine intentional states to legislatures actually plays in legal practice. According to him, this role can be best explained in the fictionalist framework.

Canale’s view seems to fit well into the broad philosophical understanding of fictionalism: his view is that even though we seem to be forced to accept, in our best theory, the truth of a certain claim X, this commitment does not need to be taken as a commitment to the literal truth of X. Moreover, we do not need to think that this commitment to the truth of X has any genuine ontological consequences: we do not need to see the things to which X refers as genuinely existing. In this particular case, Canale claims that although legal theory seems to accept claims about legislative intentions as being true, we do not need to see such claims as being literally true. Moreover, we do not have to claim that parliaments “out there” have genuine intentions. Legal theorists’ statements about legislatures believing and intending are rather made in fictional mode.

Another field in which the fictionalist approach has been developed is the theory of the firm in economics. In their classic paper, Jensen and Meckling state that “The private corporation or firm is simply one form of legal fiction which serves as a nexus for contracting relationships and which is also characterized by the existence of divisible residual claims on the assets and cash flows of the organization which can generally be sold without permission of the other contracting individuals” (Jensen & Meckling, 1976). What is important about Jensen and Meckling’s approach to the firm, which is sometimes called “the nexus of contract approach” (see, e.g., Gindis 2009 for a critical overview), is that it locates the genuine explanation of “corporate behaviour” at a “lower” level. For Jensen and Meckling, in order to understand the firm, we must look at the often-contradictory interests of stakeholders, managers, and company debt-holders.

The philosophical assumptions of the “nexus of contract” theory of the firm, as characterized by Gindis, are the following: “The first is that firms have vague boundaries or that firms are not determinate. The second is that firms are simply shorthand expressions as opposed to real entities. The third is that firms are legal fictions, which is supposed to reinforce the claim that firms are ontological fictions” (Gindis, 2009, p. 29). Although the philosophical commitments of this view are less explicit than in the case of Canale’s theory, it seems fair to say that the nexus of contract view fulfils the criteria of being a fictionalist view in a philosophical sense. On this view, although economists use talk about “firms” and “corporations” in their best theories, such talk should not be taken as expressing an ontological commitment to the existence of companies; moreover, the truth of statements about corporations (and their decisions, intentions and the like) need not be taken as literal. When constructing economic theories and descriptions of markets, it is useful to speak as if firms were singular entities that are capable of having preferences and intentions on which they can act in the market realm. Such descriptions are useful simplifications of the actual reality, in which the only objects who are genuine intentional participants in markets are individual humans. These individuals perform some of their actions under the guise of firms, which are actually nothing more than the results of contracts between individuals. However, the complexity of interactions between individuals is such that it is useful to engage in the pretense of firms being separate actors. Moreover, market phenomena descriptions which use firms as separate agents might not, in practice, be reducible to descriptions which deal only with individuals. Still, the main fictionalist idea here is that talk of ‘firms’ in theories does not incur any genuine ontological commitments.

5.1 The importance of fictionalism

It is true that there is no fictionalist consensus in either the theory of the firm or in legal theory. Still, I think that the relative popularity of the fictional approach in these disciplines should lead analytic social ontologists to treat this position much more seriously than is currently the case. It is worth noting that the commitment to fictionalism in both these areas of inquiry goes deeper than just the use of the word ‘fictional’. The idea is much more significant: basically, the view is that parliaments and firms need not be treated as real entities when we construct a theory that explains the law-making process or the economic behaviour of corporations. The basic fictionalist thought in both these approaches is that although there are no “legislators” and “firms” which exist as separate, genuine entities, it is useful to talk as if there were such things and agents.

Social ontologists of the realist and eliminativist persuasions have noted the possibility of fictionalism, but they have mostly rejected the fictionalist view. Ludwig (2017a), who is the main proponent of reductive eliminativism, has admitted that corporations function as legal fictions in the context of legal discourse, and that we often engage in a kind of pretense when attributing intentions, decisions and so on to corporations. However, his main view remains that attributions of mental states to corporations in most contexts should be treated in a reductive manner, namely as shorthand for descriptions of the actions of proxy agents, spokespersons, etc. Pettit (2014) also admits than joint-stock companies are treated as legal fiction in certain contexts. Still, he remains convinced that the most apt approach to the general issue of group mentality is the realist one.

There are several important issues that can be noted when we look at the criticisms that Ludwig and Pettit levy at fictionalism. Both theorists, somewhat unsurprisingly, consider the fictionalist stance to be at odds with the position they espouse. This is essentially right: fictionalism should be treated as opposite to both the strong realist and the reductionist projects. Both realism and reductionism assume that statements about the intentions of corporations, etc., might be true in a literal sense: for realists, they are true simpliciter; for reductionists, they are true in virtue of more basic facts; for the fictionalist, on the other hand, they should not be treated as true in the literal sense.

In order to clear up some potential misunderstandings, let us take a brief look at some of the arguments that the critics of fictionalism levy against it. Pettit (2014), for example, seems to assume that in order to deny that statements about group mental states are true simpliciter, the fictionalist must be able to perform a successful reduction of statements about group intentions to statements about intentions of individuals. For him, a powerful argument against fictionalism is the failure to equate the intentional states we attribute to groups with the aggregate states of individuals. As we cannot ‘read off’ group states from individual ones, we should see group minds as real, not as fictional phenomena.

This is, to my mind, a mistaken view on the nature of the fictionalist position. Fictionalism, in general, is not committed to any claims of reduction (see, e.g., Kalderon 2005), although certain versions of fictionalism might be compatible with them. In no way should the success of the fictionalist project with regard to the issue of group minds be seen as being hostage to the success of the reductionist project. In this respect, Ludwig might be on the right track in seeing the fictionalist position as an alternative to the reductionist one.

Moreover, Pettit (2014) claims that one reason to reject fictionalism and embrace strong realism is the fact that attributions of intentionality to groups are indispensable in theories. However, it seems incorrect to suggest that the fictionalist must claim that attributions of mentality to groups have no explanatory role to play. In some other contexts, philosophers of the fictionalist persuasion have claimed that the adoption of fictionalism allows us to retain the use of fictionalist-interpreted statements in reasonings: this is, for example, stressed by Nolan, Restall, & West (2005) in their discussion of moral fictionalism. For them, the important advantage of the fictionalist view on moral discourse is that it simultaneously allows us to deny the literal truth of moral statements whilst also using them in correct reasonings.

To sum up, the central tenets of the fictionalist position about group minds are the following. First, attributions of mentality to groups should not be treated as being literally true. The use of attributions of intentional states to institutional groups does not have to be seen as incurring any genuine ontological commitment. However, this denial of ontological commitment does not mean that attributions of mentality to groups are either reducible or eliminable.

This brief characterization of the fictionalist approach to group mentality certainly lacks some crucial philosophical detail. It might be suspected that the fictionalist pronouncements that are made by non-philosophers do not point to a single coherent metaphysical interpretation. There are many versions of metaphysical fictionalism (the pretense approach, prefix theories, and the like), and the (sometimes generic) remarks of the theorist who espouses fictionalist views on group mentality do not determine which particular version of fictionalism they propose. This, however, should not be seen as a reason to discard the fictionalist proposal. Fictionalism, even when characterized in the most general terms, should be seen as a serious contender in the debate. This means also looking into some potentially problematic aspects of fictionalism (like the issue of whether fictionalism must treat the responsibility of group agents as somehow fictional, or whether it can adequately deal with all the features of the group-agency discourse); however, to establish whether such problems are fatal for fictionalism, we must first properly describe this view.

6 Deflationism

Another possible alternative to the meta-metaphysical realist views that is worth taking into account in the present context is deflationism. In contrast to fictionalism, deflationism about group mentality is mostly a hypothetical position which has not been explicitly presented in the literature (however, deflationism is an established view in general meta-metaphysics; see, e.g., Berto & Plebani (2015), who see both fictionalism and deflationism as one of the main rivals to the orthodoxy). In my view, it is useful to construe such a hypothetical view as it would provide an interesting contrast to both meta-metaphysical realism and fictionalism concerning group minds.

The basic theoretical tools in constructing the simplified deflationary view about group minds will be borrowed from Thomasson’s approach to the existence of social groups (Thomasson, 2019), which in turn is an extension of her general ‘easy’ approach to ontology (see Thomasson 2014). It is important to note that Thomasson’s own application of deflationism to the area of social ontology has been limited to the issue of the existence of groups; she has not explicitly expressed her view on the issue of group minds. Still, to my mind, a somewhat natural extension of her position might be used to construe an interesting version of deflationism about institutional intentionality.

On Thomasson’s (2019) approach, the answer to the question of whether social groups exist is relatively straightforward: they obviously do because the application criteria of the terms we use to name groups are met. So, we might say that the Supreme Court exists because the application criteria of the term “Supreme Court” are met. Similar reasoning applies to other groups, both formal and informal. According to Thomasson, there is nothing magical about such groups coming to exist; this is because “existence”, on the deflationary approach, is a lightweight property. To say that a certain object X exists, in general, means that the criteria of application of the term “X” are met (Thomasson, 2014).

What is important about Thomasson’s deflationism about the existence of groups is that she denies that there are any essential properties of social groups: there are no necessary properties that all groups share. Second, she underscores that the questions about groups that we should be asking are of a normative nature. These include questions concerning the function of social group terms in our lives, and how they normatively structure our lives. Thus, the vital question for Thomasson is not “are there social groups?” but “what do we need social group terms for?”. This is, presumably, a consequence of her general pragmatic approach to ontological questions (see Thomasson 2014). For her, the question of whether certain Xs exist is closely connected to the question of whether we want to adopt the X-talk. The pragmatic reasons to adopt (or not to adopt) a certain linguistic framework might be of a normative or theoretical nature, as a given discourse might be useful for our practical lives or theoretical aims.

Thus, there are three important aspects of the deflationary approach to the existence of social groups: the ‘easy’ view on the question of their existence, anti-essentialism, and the focus on the normative (or, more broadly, pragmatic) import of postulating them. By focusing on these three aspects, we might try to construe a deflationary position about group mentality that would somehow mimic these claims.

The first claim would be that statements about group mentality should be seen as being true, albeit only in a deflationary sense. The idea would be that we might truly attribute mental states to group subjects, but only because institutional groups meet the folk criteria of the attribution of mental states (similarly, groups exist for the deflationist because terms that refer to groups have their application criteria fulfilled). Spelled out in more detail, the idea that mental statements should be seen in a deflationary fashion reads as follows: the fact that certain entities can be truly said to possess certain mental states stems from the fact that these entities fulfil the folk application criteria of the relevant mental terms. So, a certain agent might be said to have intentions if it fulfils the ordinary criteria of application of the term “intention”. The same can be said of beliefs: an agent can be said to have a belief that p if it meets the criteria of application of the term “believes that p” (see, e.g., Mölder 2010 and Poslajko 2022 for examples of an explicitly deflationary approach to beliefs in general).

The second component of the hypothetical deflationary approach to group beliefs is anti-essentialism, namely the idea that there are no hidden essences of group mentality that we should seek to discover. The only ‘discoveries’ we might make about the mentality of groups are of a conceptual or normative/pragmatic nature. This brings us to another important feature of the discussed position, namely its focus on the normative aspects of the issues in social ontology. On the deflationary account, the proper (and difficult) questions regarding group mentality deal with the function of ascriptions of mental states to groups. Such questions include: What is the normative and pragmatic significance of the practice of ascribing mental states to group agents? Are we willing to accept the normative consequences of this practice? Is it pragmatically advisable (from the perspective of developing scientific theories) to treat institutions as believers?

It is important to note that the notion of deflationism might be used to denote a purely meta-metaphysical view, the only commitment of which is the claim concerning the idea that existence claims are “easy” in the sense specified above; namely, that questions about existence can be answered by resorting to conceptual analysis and simple empirical observation. In a broader sense, which will be employed for the purposes of the present paper, deflationism combines the idea of the easiness of ontology with commitment to first-order realism about disputed entities. Obviously, adoption of the easy view does not lead to accepting just anything as existing as, in some cases, our common-sense criteria of treating something as existing are not met (see, e.g., Thomasson’s discussion of “wishdates” in Thomasson, 2014). However, in most cases in which philosophers treat the existence of something as being a matter of metaphysical controversy, the consequence of adopting the deflationary meta-metaphysical stance is adoption of first-order realism. This is the case of ordinary objects, of propositions, of mathematical entities, of properties, and so on. In all these cases, the folk criteria of application seem to be met, so first-order realism is a natural consequence of the easy approach.

The same seems to be true in our case: it seems that that institutional groups do meet the common-sense criteria of the attribution of mentalist terms. This is attested to by the wealth of examples in which we have no problems with saying that a given institutional group has a certain belief, desire, or intention. Such statements are abundant, both in folk discourse and in more specialist languages of legal, political, social, and economic theory. For this reason, deflationism about institutional mentality is defined in this paper as a position which includes commitment to both the easy approach to ontology and the positive answer to the question of whether institutional mental states exist.

Two important aspects of the general deflationary view are worth remembering in this context. First, the positive view concerning the existence of certain Xs, when framed in the general deflationary framework, does not have to lead to any claims about the reducibility of Xs to any “more fundamental” or “more basic” entities (see especially Thomasson 2013).

Secondly, deflationists do not claim that entities that we take to exist in a deflationary sense must play any important explanatory role. As Thomasson put it: “the deflationary meta-ontologist denies that this is a ‘deep’ and ‘explanatory’ ontological posit. The objects are not ‘posited’ to ‘explain’ what makes our claims true. Instead, the truth of the existence claim is a trivial consequence of the truth of other sentences one accepts” (Thomasson, 2013, p. 1027). This position stems from a more general idea which is central to deflationism, namely the rejection of so-called “substantial criteria of existence” (Thomasson, 2014), which include the criteria of reducibility and of causal-explanatory relevance. In Thomasson’s opinion, we should not see claims about existence as being dependent on fulfilling these criteria.

If we transpose these deflationary views onto the issue of the reality of group beliefs, we obtain the following results: the deflationist about group minds, beliefs, etc. who accepts simple first-order realism about these issues does not have to accept any claims about the possibility of reducing statements about mental states of institutional groups to statements about any more basic facts, like statements about individual minds; nor does this deflationist have to claim that group mental states are necessary explanatory theoretical posits. In the general deflationary framework, first-order realism does not entail any such claims about reduction and explanatory indispensability.

A possible worry about the deflationary view is that it might trivialize the issue of the existence of group minds. An important consequence of the deflationary approach to existence is that many of the entities/facts/properties that were traditionally considered to be metaphysically problematic are treated by the general deflationary view as existing. One interesting example is literary characters, like Gandalf. On Thomasson’s own view (Thomasson, 2003), such creatures of fiction should be seen as existing: they are abstract artifacts created by the authors of literary works. The worry is thus that the metaphysical status of groups and group minds is, on the deflationary view, similar to the status of creatures of literary fiction. In general, deflationists are often seen as promoting a sort of ‘flat’ ontology in which all sorts of entities exist in the very same deflated sense. For deflationists, however, this can be seen as an advantage, not as a problem. According to deflationists, the important philosophical questions about a certain category are not ontological: on their view, we should assume that certain ‘problematic’ entities/facts/properties exist, and we should proceed to ask other theoretical questions.

7 Map of the options

The above considerations lead us to a picture of the debate concerning the reality of institutional beliefs in which we are in a position to discern at least four options: strong realism, eliminativist-reductionism, fictionalism, and deflationism. This list is not necessarily exhaustive as other meta-metaphysical views might be possible (for example, one might think that there could be some sort of non-cognitivist approach to the issue). Still, focusing on these four options might help us get a better grasp of the central philosophical questions regarding the issue of the reality of group minds. Let us briefly recapitulate the central tenets of these theories.

The first view is strong realism, which can be characterised as the conjunction of the following commitments: meta-metaphysical realism; first-order realism about group minds; opposition to reductionism; and the assumption that attributions of mental states to groups are indispensable in explanations. The second view is eliminativist-reductionism, the central tenets of which are as follows: meta-metaphysical realism; rejection of first-order realism; reductionism; and the claim that the real explanatory work is done on the level of individual actions (this is a somewhat obvious consequence of the third claim).

The third view is fictionalism, which is characterized by the following views: denial of meta-metaphysical realism; denial of first-order realism about group minds; lack of commitment to any form of reductionism; no denial of the explanatory importance of attributions of belief to groups (although fictionalists might be committed to the idea that the “real explanations” are located on the non-group level). The fourth view is deflationism, which is committed to the following: rejection of meta-metaphysical realism; acceptance of first-order realism; no commitment to the reducibility of group mental states or to the idea that group mental states are necessary in theoretical explanations.

Once these four positions are described, it becomes clear that classifying them is not an obvious task. One way of dividing them would be to focus on the issue of their meta-metaphysical commitments: in this way, we get two meta-metaphysical realist views (strong realism and eliminative reductionism) and two meta-metaphysical anti-realist views (fictionalism and deflationism). However, we can divide them with regard to the issue of first-order realism. In this way, we get strong realism and (presumably) deflationism in the first-order realist camp, whereas first-order anti-realism is represented by eliminativist-reductionism and fictionalism.

A map of the options might be thus represented by the following Table 1:

Table 1 Map of the options in the debate about group beliefs
Full size table

8 How to choose?

In this section I will try to show how we might try to choose between the options described in the previous parts of this paper. In my opinion, the question that might reasonably be treated as the primary one in the discussion about institutional group mentality is whether ascriptions of mentality to such “subjects” can be literally true. A positive answer to the question of the possibility of ascribing mental states to groups leads to first-order realism, while a negative answer leads to first-order irrealism. However, further questions arise in both cases.

If we reject first-order realism and say that groups cannot be truly said to have mental states, then the most pressing question seems to be the issue of reduction. If we consider the project of providing a reductive explication of statements about group minds to be a viable one, then we should opt for the eliminativist-reductionist option. If, however, we do not consider reduction to be a workable option, then we should opt for the fictionalist view, which also has the advantage of allowing attributions of mentality to groups to have non-eliminable explanatory import.

If we opt for the realist view, then the question of reduction is neither here nor there. The direct consequence of realism is the claim that group mental states supervene but are not reducible to individual attitudes (see, e.g., List 2014). However, the issue which realists should treat as central is the issue of how substantial our commitment to the idea of group as believers is. If we think that the first-order realist claim reveals a genuine and substantial theoretical discovery, then one would go for the strong realist option. If, however, the commitment to the existence of group minds is taken as a minimal or deflationary one, then deflationism is the way forward.

A useful way to test the substantial character of one’s commitment to realism is by asking whether the postulated group minds are indispensable in explanations. If one believes that group mental states really do play an important role in explaining phenomena (and perhaps should be credited with the possession of distinct causal powers), then one’s commitment is substantial. If, however, we are willing to admit that we might truly say that corporations or courts have mental states but we do not consider such ascriptions to play any important role in explanations, and we do not think that group mental states can have any causal powers over and above the mental states of individuals, then our commitment is minimal. On the deflationary view, the existence of groups and of their mentality does not have to be justified by their explanatory import; this claim would most probably be rejected by realists for whom group minds, as elements of ontology, must earn their keep by proving their explanatory indispensability.

The field of the debate thus turns out to contain several interconnected but logically independent questions. It is only by answering these questions independently and trying to combine the answers into a coherent theory that we might get a fully plausible account of the alleged mentality of institutional groups.

It is not the aim of this paper to provide arguments for any of the described positions. However, it is important to observe that, in cases of both first-order realism and first-order anti-realism, opting for meta-metaphysical anti-realism might have certain theoretical advantages. Namely, in both these cases we can do without the problematic theoretical commitments that were characteristic of the meta-metaphysical realist versions of both first-order realism and anti-realism.

As has already been noted, in the case of first-order realism, the meta-metaphysical realist approach needs to embrace a commitment to some version of the claim that group minds are indispensable in explanations, and also, perhaps, to the claim that they possess causal powers, “over and above” the causal powers of individuals. But this is potentially problematic as it burdens straightforward realism with solving a version of the causal exclusion argument with regards to group mental states. Deflationism does not carry such commitments as, in the deflationary framework, claims about existence are not tied to theses about explanatory indispensability and causal relevance.

Analogously, the potentially problematic component of the meta-metaphysical realist version of first-order anti-realism (i.e., eliminativist-reductionism) was its commitment to the project of reduction: the trouble here is that these reductions, even if feasible in principle, might be deemed to be overly complex or counter-intuitive. Fictionalism, which does not adopt meta-metaphysical realism, does need to engage in this potentially futile project, and it retains the ability to deny the literal truth of group mind ascriptions.

These arguments, which invoke the notion of theoretical simplicity, do need not to be decisive. There might be important theoretical reasons for first-order realists about group minds to pursue the idea of the genuine causality of group mental states, or for anti-realists to engage in reductionist analyses. What I want to say is that the availability of meta-metaphysical anti-realist alternatives might open certain potentially interesting routes of inquiry and lessen the argumentative burdens on the disputants.

9 Conclusion

The aim of this paper is to argue that, in order to get a better grip on the issue of the reality of institutional group minds, we must allow ourselves to take into account more possible views than are usually admitted in analytic social ontology. In particular, we need to seriously consider views that reject the meta-metaphysical realistic assumptions of the previously developed views.

It needs to be admitted that neither of these views has been presented here in sufficient detail: there many subtle issues with both fictionalism and deflationism that would need to be resolved in order to claim that these views have been developed in a satisfactory way. Also, I have not presented any definite argument that would speak in favour of either of these two alternatives. I hope, however, that I have managed to convince at least some readers that treating these non-standard positions as potentially serious contenders in the debate is justified and that trying to develop these positions in more detail could lead to substantial theoretical progress. The other important conclusion from this paper is that answers to the central questions in the debate about group mentality do not necessarily come in only two tightly interconnected packages. We might and should treat the issues of the existence of institutional group minds, their reducibility to individual states, their explanatory relevance etc. as being more or less independent of one another. It seems that the proper strategy for dealing with the issue of group mentality would be to try to create a proper general view by combining the most plausible answers to these questions.