Group Lies and the Narrative Constraint

Abstract

A group is lying when it makes a statement that it believes to be untrue but wants the addressee(s) to believe. But how can we distinguish statements that the group believes to be untrue from honest group statements based on mistaken beliefs or confusion within the group? I will suggest a narrative constraint for honest group statements, made up of two components. Narrative coherence requires that a new group statement should not conflict with group knowledge on the matter, or beliefs of relevant operative subgroups, unless a coherent rationale is given. Narrative intention looks at the process of gathering new evidence on the area of expertise of the group and requires that the group position behind the statement is formed in good faith. The narrative constraint will help to distinguish group lies from more innocent erroneous statements of group beliefs when there is an internal disagreement within the group, including in cases involving spokespersons.

Introduction

When we state something that we do not mean or believe to be true, we are lying, as long as our intention is to be deceptive (as opposed to, say, engage in a mutual game of sarcastically stating the opposite).¹ Lying abuses the mechanism by which we provide reliable testimonial warrants to each other, a mechanism we need for cooperation (Shiffrin 2014). I will stipulate that to tell a lie is to make a statement that you believe to be untrue but want the addressee(s) to believe. Kelly is lying to Kwame if she says that she will meet their mother for lunch tomorrow when she has no plans to do so, just as a doctor is lying to her patient if she tells them that cigarettes have no bad health effects when she knows that they are carcinogenic.² Similarly to the individual case, a company that knows its products are detrimental to health, but claims otherwise, is lying to its customers.³ Examples of such mendacious groups are not hard to come by. It is not controversial to state that tobacco companies misled their customers about cancer and were lying to them.⁴ Fossil fuel companies have in a similar fashion engaged in lying about the effects of greenhouse gases on the climate.⁵ They have employed misinformation tactics used by the tobacco lobby to sow doubt around climate science due to financial reasons and ideological concerns about regulation (Oreskes and Conway 2010). It seems that we could say that a group is lying when it makes a statement that it believes to be untrue but wants the addressee(s) to believe.

Although accounting for lying groups might seem straightforward at first glance, how can it be established that a group makes a statement that it believes to be untrue but wants the addressee(s) to believe? After all, group members can differ greatly on their views about p, some perhaps due to ideological reasons. Cases of internal disagreements about p involving powerful spokespersons can be especially tricky. Spokespersons are proxy agents for the collective, authorized to make statements on their behalf (Ludwig 2014).⁶ Standard examples include leaders (such as CEOs) or communications staff (such as press secretaries), but a spokesperson can be anyone authorized to represent the group or a certain section of it (see also Lackey 2021: 145). To give an example from real life, consider how the then-administrator of the United States Environmental Protection Agency (EPA) announced in 2019 upon taking the job that climate change is not a top one environmental crisis (Hook and Stacey 2019).⁷ The administrator was a spokesperson for EPA and speaking as such, authorized as the leader of the institution to discuss its priorities and policies, so the claim about climate change was made by the institution, not by him as a private person. However, the claim by EPA's authorized spokesperson clashed with what EPA scientists had written in reports for years. How should we establish if a group lies in such cases? I will suggest in this paper that narrativity should factor in our judgements of whether a group is lying. I should make it clear that my discussion applies only to certain kinds of groups, namely groups that are able to form their own beliefs based on their institutional structure (List 2005: 26). These kinds of groups include corporations, governments, universities and other such highly organized collectives. It is part of the DNA of such groups that they claim expertise on some issue, be that in governance, research, manufacturing, or something else.

Although lying groups are all around us and their actions impact our lives and societies in many ways, epistemology has been mostly silent on group lies. To fill in this philosophical gap, Lackey (2018, 2020, 2021) has written about group lies and offers the following account of them.

G-LIE: A group, G, lies to B if and only if (1) G states that p to B, (2) G believes that p is false, and (3) G intends to be deceptive to B with respect to whether p in stating that p. (Lackey 2021: 186)

Group lies can be understood in terms of groups offering either coordinated, shared assertions, or authority-based assertions, meaning that the group is either making a joint statement of some form, or expressing its views via a spokesperson (Lackey 2021: 148–51). The formulation of G-LIE is clear and concise. The first condition is the default position, as there is no need to look into group lying if the group never states anything. However, although the account will work with most cases of group lying, I find that when there are large discrepancies among the members of the group on whether p, it can be hard to establish if (3) holds. This is especially the case if some of these different views are based on ideology and the people who hold dogmatic beliefs about p are in positions of power within the group. We could thus have group members that sincerely believe that p and others who sincerely believe that not-p. If the latter are authorized spokespersons, they do not necessarily intend to be deceptive, but on the group level their statement amounts to a group lie.

I will suggest that the narrative constraint can help us to distinguish lying group statements from honest albeit misguided or mistaken ones. The narrative constraint is not meant to offer an account of a group lie and it does not give us a blueprint to establish that a group is lying. Instead, it offers jointly necessary conditions for honest group statements (for highly organized groups with expertise) and can supplement Lackey's G-LIE in establishing when a group statement is dishonest rather than confused or mistaken, especially in cases where the statement is made by a spokesperson.

The narrative constraint I propose for honest group statements (i.e. for statements that the group actually believes to be true, regardless of if they are factually correct) is made up of two components. Narrative coherence implies that a new group statement should not conflict with group knowledge on the matter, or beliefs of relevant operative subgroups (if the group has any), unless a coherent rationale is given.⁸ This is because the kind of organized groups that I am discussing are subject to certain constraints when it comes to making statements regarding the knowledge it has about issues that fall within its area of expertise. This is due to the division of labour regarding group knowledge that is necessary in such groups. Group beliefs or positions do not have to be static or fully consistent with previously held group beliefs or positions, but a coherent rationale for such changes is required (see section 2.1 for more on this). Narrative intention looks at the process of gathering new evidence on the area of expertise of the group and requires that the group position behind the statement is formed in good faith. The origins of those group beliefs that are anchored in deliberatively formed group positions are subject to certain rationality constraints (section 2.2 discusses this in detail). I suggest that only when the group statement meets these two conditions can it mirror the group's actual belief. Otherwise, there is room for a group lie, i.e. deception about what the group actually believes (or for group bullshit).⁹ My proposal will help to distinguish group lies from more innocent erroneous statements of group beliefs in cases where there is an internal disagreement within the group members about some matter, including in cases involving spokespersons.

I will start with a general level discussion on what group beliefs are to provide background to those readers not already familiar with the concept. I then explore the idea of a narrative constraint for honest group statements that reflect actual group beliefs, discussing the two components of the constraint: narrative coherence and narrative intention. Towards the end of the paper, I test the narrative constraint for honest group statements by applying it to a fictional case of group mendacity, but one inspired by real-life examples of lying groups.

1. Group beliefs

This section looks at how group beliefs form – or are formed – and what is the relation of individual beliefs to that of a group. Although I will offer a stipulative account of group beliefs, the discussion is on a general level. The idea is to illustrate the issues involved to readers not already familiar with the literature on group beliefs, and to set the scene for the argument for the narrative constraint for honest group statements in the next section.

Many typical examples of group beliefs seem to comprise members deciding that the group believes that p, based on an official position that the group has arrived at through their decision-making (Lackey 2021: 34).¹⁰ While a group might come to hold proposition p, its members might not share that opinion as individuals (List and Pettit 2011). Indeed, group positions that are arrived at by group decision-making mechanisms are a product of people acting and thinking within their roles. Roles in groups tend to come with different powers and not all members have the same authority to act or to be involved in forming the collective position on a given issue. Operative members of a group are those members who have the authority to make decisions in the field in question (Tuomela 2007).¹¹ The decisions and course of action that operative members vote or otherwise opt for might not be what they would choose in their private lives. Instead, it is what they believe to be in the best interests of the group (unless they are trying to sabotage the group, are indifferent towards it, and so on). Discursive dilemmas (Pettit 2001) can further complicate the picture (I will return to these soon). So while a group might have adopted a position that p, its members might not believe it as individuals. In any case, group positions (which sometime underpin group beliefs) are not dependent upon the opinion of any given individual. For example, a single operative member refusing to accept a reason is not sufficient for the group to lack that reason (Lackey 2016: 350).

Accounts of group beliefs are sometimes divided along the lines of summative and non-summative accounts. Summative accounts (e.g. Quinton 1976) argue that if all or nearly all of the members of a group hold a belief, then aggregately the group itself holds such a belief, while non-summative, collective accounts (e.g. Gilbert 2004) deny this and demand joint acceptance of the group belief instead. These two opposites are not the only way to go. I will follow Lackey (2016) in arguing that while a group's beliefs are not reducible to its members’ beliefs, they are nonetheless constrained by them. Although a group belief does not necessarily neatly map onto the beliefs of individual group members, there is still a strong linkage between what the members-in-roles believe and what the group believes. Group beliefs are not conjured out of thin air, but are based on the beliefs of individuals within their roles. They are also shaped by the ethos of the group; that is, the goals, commitments and answers it has collectively accepted to be its view (intentional horizon) regarding the matters (realm of concern) vital to the group (Tuomela 2007; Laitinen 2014: 218). While the ethos marks the continuation of the group together with its historical and modal properties and determines the collective identity, it is not set in stone and to some degree it is always in a state of fluctuation. In my usage, the ethos (the character of the collective if you like) also includes its norms and customs, including unwritten and informal ones. Deliberations within the group can help shape the ethos through changing the group's position on some matter, but the ethos forms also organically without deliberation through the daily interactions of its members, both inside and outside the group.

I will offer a stipulative account of group beliefs, although I do not provide a full-fledged defence of it here. Rather, the idea is that the narrative constraint for honest group statements can stand on its own and can be applied to the preferred account of group beliefs.¹² Still, offering this stipulative account of group beliefs should make it easier for the reader to follow the discussion on groups who make lying statements in the following sections. I suggest that a group cannot have a belief that p if none of its members have such a belief or a belief close enough to accept that p.¹³ Instead, a significant proportion of the operative members (with the ratio depending on the context) should believe – within their roles – that p (or a belief close enough to accept that p) in order for the group to have a belief that p. ¹⁴

Group belief

Group G believes that p iff a significant proportion of the operative members of G pertaining to that matter hold individual beliefs-as-members that p (or close enough beliefs to accept that p), and such a belief belongs within the remit of G's ethos.

The distinction between individual beliefs and what I call individual beliefs-as-members is the key to understanding why the account proposed does not represent either of the furthest ends of the group belief debate (i.e. the simple summative account or the plural subject account of collective belief, to use terms from Gilbert (2004), or the deflationary or inflationary accounts, to use terms from Lackey (2016)). When an individual holds or forms a belief as a member of a group, the ethos of the group is already involved. Members look at the issue from the viewpoint of the collective and their individual deliberation is affected by their role and the ethos of the collective, just as it is when group deliberation is taking place.

Think of a customer services representative, working alone at the information desk of a department store, having to take on-the-spot decisions on behalf of the department store when reacting to customer enquiries (and being authorized to do so within a certain limited remit). If they are acting within their role requirements, their answers are not based on their personal individual beliefs, but will instead try to mirror the policies and the ethos of the department store. When it comes to group deliberation, which represents the more robust type of group belief formation, the group members are still thinking within their roles and adopting the point of view of the group. However, their individual beliefs-as-members have the additional possibility of shaping the individual beliefs-as-members of the other members too, thus affecting the group belief and even the group ethos.

Beliefs-as-members must fall within the ethos of the group. Even if all of the members of a group individually believe that p, even for reasons that are similar or complementary, the group should not be attributed with the belief that p if it is something that falls outside the ethos of the group. Let's say that after watching Chewing Gum and I May Destroy You, all the group members individually believe that Michaela Coel is one of the most talented young screenwriters around. Member A believes this is because Coel is such a versatile writer, mastering comedy and drama alike. Member B thinks it is because she knows how to write believable and multi-faceted characters, while member C's reason for the belief is that Coel's stories are not formulaic but break new ground and keep the audience on their toes. The group in this example is not the BBC or some television production company, however, but a small company that manufactures stationery. While questions about ethical supply chains or varieties of recycled paper fall safely within the ethos of this group, questions about excellent screenwriting do not. So even though the individual operative members all believe that p, the stationery manufacturer does not hold any group beliefs about the matter. This is because the beliefs about Michaela Coel are individual beliefs of members, not individual beliefs-as-members.¹⁵

Note that I did not argue that the group cannot be attributed with the belief that p just because it is something that they have not deliberated as a group. This is because I think a group belief can be attributed correctly to a group even when the members have never discussed p directly. Here my account differs from the collective acceptance accounts of group beliefs, as the members of the group do not have to openly express their readiness to establish the belief that p as the group's belief (cf. Gilbert 2004). Perhaps it goes without saying at the stationery supplier that the products would never feature any controversial images, and this can be established by looking at the product lines over the years. “Cute is better than political” could thus be a belief attributed to the group, even though it has never been openly discussed in meetings. Rather, it is part of the ethos that members almost unconsciously adopt as they get acquainted with the company's culture. Another example could be a misogynistic group that believes that women are less capable than men, but who have never discussed the issue openly. Despite this, I think it is correct to assign a group belief that women are less capable than men to the group if a significant proportion of the operative members think so within their roles and such a belief belongs within the remit of the group's ethos.¹⁶ A group belief can be formed unconsciously or unintentionally through the interaction of social norms within the group and the kind of members the group has. We could call this kind of belief dispositional group belief, meaning that it is a belief that belongs within the group's ethos, but it is not actively being considered by the group (as opposed to occurrent group belief, beliefs that are actively considered by the group).¹⁷ A group's ethos is revealed not just through what it states to be its position, but also through what it does, through its actions.

A dispositional group belief can be made visible by bringing the issue up with or within the group, by external or internal agents. This would be the case if the misogynistic group would be called to account for its misogynistic tendencies, for example. Dispositional group beliefs can be changed through group deliberation, after which it becomes an occurrent group belief, i.e. a belief that is actively considered by the group. With these latter kind of group beliefs there is a deliberately formed group position regarding p that anchors the occurrent group belief about p. I will suggest later that the genealogy of the group belief, if you like, matters for narrative intention, where it places certain rationality constraints on occurrent group beliefs, although not on dispositional ones.

While group beliefs cannot be entirely removed from what the group's members believe within their roles, this does not apply to group positions. Instead, the members might decide to adopt a group position that does not really reflect the opinion of the members due to some pragmatic consideration, like the need to reach a compromise, or the restrictions imposed by voting rules. Discursive dilemmas discussed in the literature show that even if all members of a group hold judgements that are individually consistent, the group judgement reached through aggregation of votes does not have to be consistent with these individual votes (e.g. Pettit 2001; List 2005; List and Pettit 2011). Discursive dilemmas can also consist of some rule, resulting in the group deciding on a sub-optimal choice. For example, a group can decide to hire a candidate that no member of the hiring committee considers to be the best candidate. I find that while such a result might be the group's decision, it is not what the group believes to be for the best. It could be, if the group thinks that there is no ideal procedure for coming up with a decision and they are happy to have merely settled on a procedure. This would fall under individual beliefs-as-members that are close enough to accept p. However, it could be that the rule is externally imposed, from people higher up in the organizational hierarchy, for example, and the group members are bound by the rules, regardless of which procedure they themselves would prefer. In such circumstances, a subgroup's decision to hire candidate T does not automatically mean that the subgroup believes that T is the best candidate. The individual members might all find him to be a suboptimal choice. Rather, the belief might take the form of: “We believe that K or J is the best candidate, but due to the externally imposed rules, the group will vote to hire T instead.” While the subgroup believes that candidate T should be hired, their group belief does not need to include some further component of believing that T is the best candidate. In other words, the fact that none of the individual members think that T is the best candidate blocks the possibility that the group thinks that T is the best candidate.¹⁸ Again, they might think that T is, for example, the most suitable candidate under the rules that are in place, but it would be misleading to describe it as the subgroup's belief that T is the best candidate for the job.¹⁹ The group members might have different reasons for believing that p, but as long as a significant proportion of the operative members believe that p within their roles, the group can also believe that p.

As group beliefs are constrained by the beliefs of their members, changes in membership can affect the group's beliefs. Political parties with a radically new leadership are probably the clearest examples of this. However, as most members affect the ethos of a group in myriad small ways, the changes are often more gradual and less noticeable. All in all, group beliefs are never set in stone and the ethos of a group is always in a state of flux to some degree. Changes in a group's membership can also change the justificatory status of its beliefs (Lackey 2016: 391). Although allowing for change is required for any account of group belief to be realistic, group beliefs cannot change on a whim. For something to count as a belief, there has to be a certain level of stability to the belief. Or as Lackey (2021: 42) puts it, a group belief is incompatible with judgement fragility, where the group deliberations are likely to yield different answers each time, even with the same available evidence. She likens this to an untrustworthy person, who changes their mind about a question every few minutes.²⁰

2. Narrative coherence and narrative intention

In this section, I will elaborate on the idea of a narrative constraint for honest group statements (for highly organized groups with expertise), looking at the two components of the constraint. I begin with narrative coherence before moving on to narrative intention. I will also make some remarks about group knowledge as it is an important element of narrative coherence. The narrative constraint should help us to establish when a group is lying even in cases where some of its operative members sincerely believe in falsehoods due to ideology.

A group could also be narratively confused. Narrative confusion is not an example of a group lie. Rather, it is about a group agent who is getting pulled in different directions by their members, where the ethos of the group is weak or has not yet fully formed, or where the group goals and beliefs are inconsistent. A group like this can make statements that are not true in relation to how the world is, yet these statements are not meant to be deceptive, but are the product of haphazard group knowledge and unformed group dynamics. The group can make narratively confusing statements, not because it aims to be deceptive, but because it does not really have clear aims as a group yet. An example could be a newly formed cooperative or a political party that has not yet had time to really come together as a group. This kind of confusion is not an example of a group lie. I will set aside cases of narrative confusion as they mostly apply to new and emerging groups and not to the kind of groups that I am using as my examples.

Lying should also be distinguished from honest mistakes. This can be done by observing the severity of the incoherence or the repetition of the false claim. The severity of the incoherence concerns how wide the gap is between what the group knew and what it said. Repetition of the false claim refers to – unsurprisingly – how often something is said. A few times could indicate an honest mistake (or narrative confusion), but a repeated false message is a lie. Lying group statements purposefully do not reflect what the group knows about the issue(s), whereas misguided group statements suggest either group confusion or lack of knowledge about the issue (by the spokesperson or the group).²¹ Naturally there will be cases where it is harder to establish if a group assertion is a mistake or a lie, especially if made by a spokesperson, but the narrative coherence introduced next should help with these.

2.1. Narrative coherence

Narrative coherence does not mean that group beliefs must be static or neatly converge with previous group beliefs: they do not have to be fully consistent. There can be changes, even radical ones, but a coherent rationale for such changes must be given.²² This should hold even when a group's membership changes and the answers it has collectively accepted (explicitly or implicitly) as the group's view on a given issue change as a result.

Narrative coherence

To reflect an actual group belief, a statement by a group cannot conflict with earlier statements or group knowledge on the matter, or beliefs of relevant operative subgroups (if the group has any), unless a coherent rationale is given.

A coherent rationale can be the availability of new evidence, like when a university announces that it will abandon its old research approach towards some disease, in its place pursuing a new line of research due to obtaining game-changing results. New evidence can also appear from outside the group, like when another group has made a breakthrough or previously unavailable evidence has surfaced in another way. New evidence can also be more mundane than this, like when a company receives market analysis that shows its competitors are making a lot of money on a product line in an area it has not yet branched out to and, as a result, the company decides to branch out to this area too, although it previously resisted doing so. Instead of new evidence, a coherent rationale can also be provided by previously marginalized evidence gaining a stronger foothold in society, like when testimony is heard from groups who used to be sidelined in mainstream discussions. A coherent rationale can also allow for backtracking on earlier statements and projects if the group admits to a past faulty belief, like when a government apologizes for the state's past treatment of minorities as it changes its policies.

Something is not automatically a lie if (in groups where tasks are divided along subgroups) the operatively relevant subgroups hold views that differ from the top-level group belief. Let's say that the product development team at the stationery company has formed a group belief that x is the best course of action in relation to how to approach a new market, whereas the marketing team has formed a group belief that y is the best approach. Both group beliefs are formed in light of the point of view and evidence available to that subgroup, based in part on existing group knowledge. The board members and executives of the stationery manufacturer decide to go with plan z, an amalgamation of approaches x and y. Although its subgroups hold different group beliefs about the matter, the stationery manufacturer's belief that z is the best course of action is the group's belief as long as the operative members in relation to that decision can provide a coherent rationale for why they believe that z is the best course of action (or x or y, if they decide to go with them instead). Still, a group belief cannot sidestep the operative subgroups’ beliefs pertaining to the matter unless a coherent narrative for such an exclusion exists. In addition, the stationery manufacturer's belief that z is the best course of action can be an actual group belief (instead of a lie) only as long as z does not contain a denial of knowledge that the group has.

In order to highlight the importance of the last point, it is worth pausing for a moment to consider what group knowledge consists of. An internal division of cognitive labour often exists within groups (Bird 2014). Knowledge does not need to be evenly distributed within groups to still count as group knowledge, but the way it has been arrived at and by whom matters (Hormio 2020). For a group to have knowledge about x, not every member or even the majority of members needs to have knowledge about x. My conception of group knowledge is thus not simply summative, although it of course allows that some group knowledge can be shared among all the members. However, requiring all knowledge to be shared knowledge within a group would be implausible, especially whenever we are talking about larger groups. Instead, what is required is that enough of those who are active in the areas requiring this knowledge possess it or could access it, namely the operative members in relation to that issue. This kind of group knowledge is operating knowledge (Hormio 2020). Operating knowledge is compartmentalized, known only by the operative members. It is often specialized knowledge: you need a certain level of expertise to be selected for an operative role within a group. While the knowledge is usually shared among the operative members, it is not shared knowledge among the group as a whole. Note though, that operating knowledge does not necessarily have to be shared among the operative members, either. Sometimes it is not traceable to knowledge of any current members, but is instead encompassed in files, for example, that are accessible to operative members (Hormio 2020).²³ Sometimes operating knowledge resembles distributed cognition, where various distinct subtasks contribute to an overall task, which is the case with complex long-term scientific experiments, such as those conducted at CERN's particle physics laboratory (see Bird 2014).

What is crucial is that to count as operating group knowledge, the knowledge must be attached to the relevant roles and appropriate lines of communication (Hormio 2020). To give an example, a choir with 50 members has a number of upcoming concerts for the spring. Only three people in the choir know about the provisional concert dates for the autumn as they are the members tasked with marketing the choir and booking concerts. The provisional schedule for the autumn concerts is not shared knowledge within the group, but it is nonetheless group knowledge in the operational sense. Still, in order to ascribe operating knowledge of a certain matter to a group, that knowledge must fall in line with the group's structure regarding roles and lines of communication. In the case of the choir, the choir leaders know that the operative members have a provisional schedule planned for the autumn because the marketing team has updated them on the matter. The knowledge about a provisional concert schedule for the autumn is operating knowledge for the choir, although it is not yet shared knowledge. This holds regardless of whether the choir leaders know about the provisional dates, or if they only know that there is such a provisional schedule.

For honest group statements there needs to be coherence between what the group states as its beliefs and operating knowledge of its subgroups. An honest group statement can of course be factually incorrect (i.e. the group can have false beliefs), but it can only be an honest reflection of what the group believes if it is not divorced from the operating knowledge the group has on the issue at hand. This applies also to possible ideological concerns of some members, which cannot override evidence that the group already possesses, including when group assertations are made by spokespersons. I agree with Lackey (2021) that spokespersons are not parrots with a ready-made script but can assert on behalf of the group without consulting the group first, as long as they have the authority to convey information on the matter and they act in virtue of this authority. The spokesperson could be a group member and might even already wield power within the group due to their position (like the CEO), or they could just be authorized to speak on behalf of the group, without necessarily having any powerful role within the group, like a communications officer. They could also be external to the group and hired to represent them, such as a PR person or a lawyer, so spokespersons do not necessarily have to be group members.

What matters for my argument is that even when they have the authority to speak on the subject matter, assertions by spokespersons need to be made within the constraint of narrative coherence if they are to reflect actual beliefs of the group. I thus propose that the narrative constraint – in the form of narrative coherence – applies to the autonomy of the spokesperson to make honest assertions on behalf of the group, and this links to the knowledge that the group has on the issue. The idea is not to try to rule out spokespersons misrepresenting actual group beliefs, as this can happen for a variety of reasons (inept spokesperson, misunderstanding etc.). Rather, the idea is to show that statements by spokespersons that knowingly set aside group knowledge (i.e. violate the narrative coherence) amount to the group intentionally lying.

Recall the EPA administrator announcing in 2019 that climate change is not a top one environmental crisis. Such a claim is incoherent with the knowledge that EPA has amassed about climate change over the years and what it has previously said about the issue in its statements and reports. As there is no new scientific evidence about climate change being less of a threat, or about climate models being less accurate or less alarming (which could explain why it should no longer be a top priority for EPA), the assertion by the spokesperson falls foul of narrative coherence. Lackey (2021: 151–2) does discuss rogue spokespersons, who assert p on behalf of the group without the authority to do so (or without doing so in virtue of this authority), distinguishing these from bad spokespersons, who are incompetent or negligent. However, the EPA administrator does not fall into the category of rogue spokespersons: he had the authority to represent EPA and make the assertion that he made. Still, the assertion by EPA was a lie, regardless of the personal beliefs and ideologies of its leader (i.e. whether the spokesperson truly believed it himself), because it fails to satisfy narrative coherence. Claims by spokesperson should not be incoherent with group knowledge if they are meant to represent actual group beliefs, even when the spokesperson is authorized to speak without consulting the group.

Now, Lackey (2021: 153) acknowledges the worry that according to her view, a group asserts something through a spokesperson, even if every member protests the statement. Still, she thinks this is a correct result and shows why we should choose our representatives carefully. I agree with Lackey that the group asserts p if the spokesperson is properly authorized to do so, even if every member protests the statement. However, I want to make the further argument that an assertion like the one in the EPA example is not an honest assertion of a group belief, but instead a case of a lying statement, as it falls foul of narrative coherence. It is made by a proxy agent who is authorized to speak about p and represent the group's views on p, therefore making it a statement by the group agent. However, it is not an honest statement as it misrepresents the group's actual belief about p. Narrative coherence is especially important for institutions and other organized groups that position themselves as experts in their field and bank on their reputations. Although EPA administrators are appointed by democratically elected politicians and therefore reflect the political winds (to some degree at least), in leading the institution and in making statements on behalf of it, they cannot set aside previous research conducted by the EPA if they are to maintain the credibility of the institution.

The case also illustrates that sometimes group members have little or no say in who is representing them as spokespersons, so the group as a group might not be able to choose its representatives carefully. Government agencies are usually led by people appointed by politicians and although they are often qualified for the job, this needs not be the case. The phenomenon of not being able to choose your spokesperson is not limited to government agencies and is currently the case with many workplaces. We often do not get to choose our bosses or who speaks on behalf of the company, institution, association etc. we belong to, unless the workplace is organized in a democratic manner. Many groups and institutions fall somewhere between democratic and authoritarian, and it matters a lot whether the spokesperson is asserting things that fall in line with the group's actual beliefs.

I think there is a need to also separate group lies made by spokespersons (without the backing of the wider membership) from misleading statements made by bad spokespersons. Bad spokespersons, who are either incompetent or negligent, can be reprimanded by the group and the group could retract what they asserted about p (Lackey 2021: 152–3). It seems highly unlikely that this is what happened with the EPA case, especially since the person doing the asserting on behalf of the group was the most senior person in EPA at the time (and the people holding political power were presumably satisfied with such a stance). But perhaps we could argue that the administrator was a bad spokesperson regardless of not being officially treated as such by EPA, he just had enough power to get away with it. Such a conclusion leaves open the possibility that the statement was just an incompetent misrepresentation of an actual group belief by a bad spokesperson. I want to deny any vagueness about it and argue that because the assertion was clearly narratively incoherent with earlier EPA statements, and its current operating group knowledge about climate change, it is a case of a group lie, regardless of the intentions of the spokesperson (or the lack of intention by the other group members to engage in deception, or in making such a statement in the first place). The intentional deception in this case thus takes place at the group level: it is in the incoherence of the new and earlier statements, and the incoherence of the new statement and the current operating knowledge.

Recall that in addition to coherence with regards to knowledge, the narrative coherence requires that an honest group statement should not conflict with beliefs of relevant operative subgroups, unless a coherent rationale is given. It seems that there is something amiss if we allow assertions by powerful spokespersons to shape the beliefs of a group with no constraints (not that I am suggesting that Lackey would endorse such a view, but only pointing out that this possibility seems not to be straightforwardly ruled out in her account). Ludwig (2017, 2020) has suggested that the sincerity of group assertions is decided based on whether the group members are jointly committed to acting in conformity to the assertion (and that talk of group beliefs and acceptance is just a shorthand for this). I think this is along the right tracks, as what the (operative) members believe constrain what can be in an honest way asserted by a spokesperson. I do not mean to suggest that everyone has to be aware of the issue and the group's answer to it for it to count as the group's position on something. No mutual awareness of acceptance is required either. The operative members adopting something as the group's position or view within their jurisdiction without other members knowing about this suffices for acceptance in my account, as long as if the other members were to hear about the position, they would not feel compelled to protest against it. There is a level of tolerance of a compromise that members must have in most group settings, at least within large groups.

2.2. Narrative intention

I now turn now to the second component of the narrative constraint: intention.

Narrative intention

The group position on whether p should be formed in good faith based on the best available evidence if it pertains to an area of expertise for the group.

Recall how with occurrent group beliefs (i.e. beliefs that are actively considered by the group), there is a deliberately formed group position regarding p that anchors the group belief about p. The narrative intention does not apply to dispositional group beliefs, as these could have arisen without deliberation, so there needs to be no deliberately formed group position that underpins such beliefs. Both types of group beliefs can be irrational, dogmatic, or unfounded, but only those group beliefs, which have been actively considered by the group at some point, should be evaluated through the lens of the narrative intention. These beliefs might not be occurrent group beliefs as such anymore, but they have a group position underpinning them that was at some point deliberately formed. In other words, their genealogy includes them being actively considered at some point by the group. So even if the group belief is no longer actively considered by the group, and has not been for a while, the group belief is underpinned by a deliberately formed group position. It is these kinds of group beliefs that I will discuss in this section, arguing that the narrative intention can help us to differentiate group lies from statements of group beliefs that are irrational, dogmatic, or unfounded, for example, but do not amount to the group lying.

Narrative intention does not require that the group belief is justified. After all, honest group statements can be factually incorrect and/or based on poor evidence, but as long as they reflect what the group actually believes to be true, they are not group lies. The genealogy of the group belief matters for narrative intention, as it places certain rationality constraints on those group beliefs that are anchored in deliberatively formed group positions.

Narrative intention requires a certain level of epistemic curiosity and openness from the group to engage with the best available evidence and different views on an issue that pertains to an area it claims expertise in. If such willingness to properly engage with evidence and views that might challenge the group's existing views is missing from the process of gathering evidence, then the group position is formed with the narrative intention of upholding the status quo. This goes against the idea of forming group positions in good faith, where deliberations should be based on the best available evidence, not on dogma or wishful thinking. It is not enough for narrative intention that the group position is based on the evidence brought before the group at some specific meeting. Instead, the evidence has to be the best evidence that is available to the group, meaning that it would be feasible for them to obtain it should they wish to do so.

The process of forming group positions involves the group members discussing available evidence and debating their options. However, this process can easily be manipulated. The group members could, for example, order a report from an outside consultant that has a very particular frame of questioning. They could also decide in advance of a meeting to only admit evidence that accords with what they would want the group belief to be (Lackey 2016). Both are examples of selective evidence-gathering without epistemic curiosity or proper care for the best available evidence. Indeed, the group can be engaged in self-deception or even lying when it does so. Narrative intention rules out group positions that underpin group beliefs that are formed through manipulative means. Although the misleadingly selective choice of evidence can be made intentionally, it can also happen semi-unconsciously through strong biases. Examples will be easy to find in recruitment committees, for instance, that are unconsciously influenced by the names of the candidates to omit most non-white applicants from the shortlist. This is why some form of care to make sure that group decisions are made by a representative body is also required for narrative intention.

At times, a group might not know about the availability of particular evidence, and at other times it may have genuine blind spots. In these cases, the group belief can meet the requirement of narrative intention, even though it is not based on the best available evidence. Think of a group that has selected a treatment for its members and believes that this is the best possible treatment, without being aware that a new, better alternative has recently been developed. The group had conducted reasonable research into the available alternatives, but it simply did not happen to come across information about the new treatment, as news of this development had not yet become widespread, due to the developer having very limited means to market their invention. The group belief that p was the best available treatment, even when a better treatment had been developed, meets the requirement of narrative intention as it was formed in good faith. That said, such a group should be willing to take into account additional evidence if and when someone challenges it on its false belief. If the problem seems to be implicit biases or the lack of epistemically robust viewpoints on an issue, the group should take steps to diversify its operative membership to try to avoid coming up with such biased or issue-blind beliefs in the future.

3. Before we knew

In this section, I will discuss group lying in the climate change context to test the applicability of the idea of a narrative constraint for honest group statements. Many corporations that depend on fossil fuels have engaged in lobbying to prevent regulation at local, national and international levels (Dunlap and McCright 2015).²⁴ Their own scientists understood the dangers, and produced models as early as the 1980s, which showed that immediate action should be taken. Instead, the companies chose the opposite course and invested heavily in lobbying regulators to adopt a wait and see approach, some of them at least partly for seemingly ideologically reasons (Oreskes and Conway 2010).

My example is a fictional fossil fuel corporation F, who publicly denies what its own experts have told it about the gravity of the threat posed by climate change and the reliability of climate science. The narrative constraint should help us to see if the group is lying when its operative members are internally disagreeing about what to do. This is even if some members are guided by ideology over facts, meaning that they – as individuals – could sincerely believe in falsehoods. Adopting the formulation of G-LIE from Lackey (2021) and applying it to F, the criteria for a group lie would be as follows:

F lies to its consumers if and only if (1) F states to its consumers that p (climate change science is unreliable), (2) F believes that p is false, and (3) F intends to be deceptive towards its consumers with respect to whether p by stating that p.

Following from the earlier section on group beliefs, I suggest that F believes that p iff a significant proportion of the operative members of F pertaining to that matter hold individual beliefs-as-members that p (or close enough beliefs to accept that p), and such a belief belongs within the remit of F's ethos. I argued that while a group's beliefs are not reducible to its members’ beliefs, they are nonetheless constrained by them (I will return to this soon).

Let us say that F places an ad in a newspaper, stating that climate science is poorly understood and that there is no proof that the human use of fossil fuels has had an impact on the average global temperatures. This is despite F's own scientists having written an internal report, which confirms the science, acknowledges the potential serious implications for the world and urges F to consider how it should play its part. In the context of this example, the third condition seems to follow from the second one holding. After all, when you pay for advertising space, you have a message that you want to convey to the audience of that media. Moreover, if you publicly state something that you believe to be false, and do so in persuasive marketing language, it would be hard to deny that the intention is to be deceptive.

However, this is too fast. There could be two different versions of F. In the first, the intention is to be deceptive about climate science because of concerns over short-term profits. The members of the board of directors of F believe the conclusion of the internal report, but decide that F's official position is that climate science is uncertain and any climate legislation would be harmful, purely in order to protect F's short-term financial interests. Here the ad amounts to an obvious group lie. In another version of F, some of the operative members truly believe that climate science is not to be trusted. In what follows, I will discuss only the latter version of F, where its operative members are strongly ideologically influenced.

Ideological Fossil Fuel Company

Fossil fuel company F places an advertisement in a prominent newspaper. In the ad, F claims that climate science is poorly understood and that there is no proof that the human use of fossil fuels has had an impact on the average global temperatures. This is despite F's own scientists having written an internal report, which confirms the science, acknowledges the potential serious implications for the world and urges F to consider how it should play its part. The members of the board of directors of F genuinely do not believe the conclusion of the internal report due to ideological reasons, and decide that F's official position is that climate science is uncertain and any climate legislation would be harmful.

For board members of this F, climate change is just a socialist ploy to introduce regulation against free markets, so it is less clear that they intend to be deceptive when they put out the statement in the advert. Rather, F seems to be only expressing the actual ideological group belief of its board members. I will suggest that despite the genuinely held beliefs of F's operative members, F is lying when it puts out the advertisement. The statement is a group lie because it fails the narrative constraint for honest group statements.

Returning to the second condition of G-LIE (F believes that p is false), and my stipulated account of group beliefs, one might worry that reliance on the beliefs of the members provides leeway for groups to engage in “alternative facts” that best fit their existing ethos by choosing only ideologically like-minded people as members. However, even if a significant proportion of F's members believe that climate science is akin to a conspiracy theory, it does not mean that this would translate into such a belief for F because the members would have to be operative members in relation to the issue. The operative members regarding F's beliefs about the reliability of climate science are F's own scientists, the experts F has tasked with compiling the internal memo about the subject. Their report confirmed the science, acknowledged the potential serious implications for the world, and urged F to consider how it should play its part. As the company scientists are the operative members in F when it comes to assessing the reliability of climate science, F believes that climate change science is reliable. If these scientists ignored the best available evidence, they would fall foul of the narrative intention requirement for forming group beliefs. If F hired a bunch of people without the necessary skills and knowledge to act as its company scientists in relation to climate change, namely people who are not experts in sciences directly related to climate science, then F itself would not meet the narrative intention requirement.

That said, the company scientists are operative members only in relation to F's beliefs about the reliability of climate science. F's executives and board members are the operative members in relation to the belief of what F should do in light of this evidence about the impact of their products. Let us call them operative members in relation to F's response to climate science and stipulate that a significant proportion of them believe that climate science is a hoax. This is not an implausible scenario, particularly over three decades ago when decisions about misinformation campaigns were first made. There can be many ideological obstacles involved in accepting climate science and the need for environmental regulation.²⁵ Viewed from the standpoint of a board member of a fossil fuel company, the emerging climate science must have been alarming not only because of the changes projected to occur to global average temperatures, but also because such science made it clear that time would soon be up for fossil fuels. Unless a company is willing to change its course very rapidly, such news can be almost too painful for its members to process, even when it comes from the company's own scientists (or perhaps especially then). It is therefore not hard to imagine how this could result in denials and selective framing in group decision-making. So, it seems that the group could believe that p, meaning that the second condition of a group lie (F believes that p is false) would not hold. This would be too hasty a conclusion to draw, however.

When a group has internal discussions about an emerging issue and is looking to decide what should be done about it, the group position is the product of group deliberation. If the group's position on an issue is based on biases and wishful thinking, instead of the operative member's deliberations and the best available evidence, the position amounts to a group desire, rather than a belief. After all, as Anscombe (1963) pointed out, a belief has a mind-to-world fit: we try to match our beliefs to how the world really is and correct them with the availability of new evidence. A desire has a world-to-mind fit instead: it is about how we would like the world to be, how we would like the reality to match our desire. If F's operative members in relation to its response to climate science (i.e. executives and board members) choose to ignore the evidence that their own scientists have provided them with, and let their biases guide them instead, the resulting group belief based on a group position is more akin to a group desire than a group belief.

I argued earlier that a group belief born out of conflicting beliefs of different subgroups can be the group's actual belief as long as the narrative is coherent, but that a group belief cannot sidestep the beliefs or knowledge of the operative subgroups pertaining to the matter unless a coherent narrative for such an exclusion exists. When fossil fuel company F claims that climate science is unreliable – even when its own group of experts has internally concluded in its confidential report that the science is reliable – the lie is in the discrepancy between what the company knows internally through its experts and what it says externally. As the paid advertisement states that climate science is unreliable, the statement is a group lie. The only way the statement about unreliable climate science would not fall foul of the narrative coherence condition would be if F had another, reliable source of information that concluded that climate change is a hoax, or the science unreliable at best. Such a dissenting opinion cannot come from just anyone, however. As I said earlier, if F hired people without the skills and knowledge to act as experts in relation to climate change, then F would not meet the narrative intention requirement. Hiring a consultancy that provides ideologically fitting answers would not get F off the hook either. This would also be the case if F decided to fund only the handful of climate scientists that are critical of anthropogenic climate change. Deciding to rely on such niche experts and ignore the scientific consensus falls foul of the narrative intention requirement.

After surveying the evidence, F's scientists concluded that the science is real. With regards to the advertisement, there is thus a problem with narrative coherence even if F's executives and board members have always believed that fossil fuels are good for the world. While the new group belief does not necessarily contradict earlier group beliefs about what the group should do, there is incoherence between what F knows about the science through its experts, namely the knowledge of F's operative subgroup, and what it states publicly about the science. It would be another matter if F stated in its adverts that climate science is reliable, but tried to argue that no immediate action is required due to economic concerns, for example. But this is not the case in my example. Thus F's statement in the advertisement does not meet the narrative coherence condition.

The narrative intention in this case also points to deliberate lying as opposed to accidental misleading statement. Group positions should be formed in good faith based on the best available evidence. I argued earlier that narrative intention requires a certain level of epistemic curiosity and openness from the group to engage with different views and evidence on an issue. In this case, it seems that such curiosity existed: F tasked its own competent scientists with compiling a report about climate science. However, it has not engaged with this evidence in good faith, but instead dismissed it or perhaps decided to hide it. If such willingness to engage with evidence is missing, the group belief is not formed in good faith, as deliberations should be based on the best available evidence.

The public statements made by a group can of course be inconsistent with its operating knowledge, which seems to be the case with the fossil fuel companies. It might even be that the statements are inconsistent with the group's shared knowledge, but that is beside the point. What matters for the purposes of my argument is that operating knowledge about x is enough to give rise to a group belief about x as long as that knowledge is attached to the appropriate lines of communication. In the case of F, the scientists compiled their report and shared it with the board members. The operating knowledge that F's scientists have about climate change science was thus attached to appropriate lines of communication (a report to the board members). To disregard such knowledge when making decisions about what the group should do fails the test of narrative coherence. Hence, the group is lying.

4. Conclusion

I have addressed the phenomenon of group lies in this paper and suggested that narrativity should matter for our judgements of whether a group is lying or not. More precisely, the narrative constraint for honest group statements can help us to identify group lies (in highly organized groups with expertise) even when some operative members sincerely believe in falsehoods. The narrative constraint offers jointly necessary conditions for honest group statements (i.e. statements that the group actually believes to be true). Narrative coherence implies that a new group statement should not conflict with group knowledge on the matter, or beliefs of relevant operative subgroups (if the group has any), unless a coherent rationale is given. This applies even when group statements are made by spokespersons authorized to speak on behalf of the group. Knowledge can be ascribed to a group as long as it is attached to the group's structure and lines of communication, even if it is not shared knowledge within the group. The beliefs and knowledge of the relevant operative members cannot be sidestepped by ideological beliefs of even powerful members. Narrative intention requires that the group position behind the statement is formed in good faith based on the best available evidence if it pertains to an area of expertise for the group. While it does not require that the group belief is justified, the genealogy of the group belief places certain rationality constraints on those group beliefs that are anchored in deliberatively formed group positions. Narrative intention requires epistemic curiosity and openness from the group and rules out forming group positions that underpin group beliefs through manipulative means.

There is considerable variety within group lies that falls outside the scope of this paper. The normative “badness” of group lies can differ in the same way as it can in respect of individual lies. A group could be telling harmless white lies to protect someone, or it could be telling manipulative, harmful lies to try to avoid responsibility for its actions or to try to stage a coup, just to give some examples. While the group has issued a statement that amounts to a lie, only a handful of people could be behind it, or the whole group could be accountable for the lie. The responsibility of individual members will differ according to their role within their group and their knowledge of and participation in the lie. The group could also be lying to its members, with just some of the operative members in on the real goals of the group, like in a cult where the regular members are misled about the true goings-on. The group could also pretend to believe in something, as in the case of the Flying Spaghetti Monster, but not with the intention to be deceptive, but rather to make a political statement.²⁶ These kinds of “as if” group beliefs could make it hard to determine when the group is lying, especially if some of the members believe the proposition and others only pretend to do so. These are interesting issues, but the goal of this paper has not been to offer a comprehensive account of group lying that takes all such questions into account. Instead, the goal has been to suggest a narrative constraint for honest group statements that can help us to identify group lies even when there is significant divergence on the issue among the members.²⁷