On Gregariousness*

There seems to be a difference between drinking coffee alone at home and drinking coffee in a café. Yet, drinking coffee in a café is not a joint action. It is an individual action done in a social environment. The café, with each person minding their own business next to others, is what I call a gregarious state of affairs. Gregariousness refers to the warmth of the social world. It is the difference between studying alone at home and studying in the library. This light form of sociality is precisely what we were deprived of during the coronavirus lockdowns. Gregariousness cannot be explained as interaction or coordination, and neither can it be grasped solely as a normative aspect of the environment. This is why gregariousness cannot be explained using the concepts of strategic equilibrium, shared planning agency, joint commitment, we-intention, or second-person standpoint. In this paper, I will also provide a prospective theory of gregariousness. The aim of this paper is not to provide a definitive theory of gregariousness, nor to demonstrate that other theories of joint action are incorrect, but rather to draw attention to this aspect of the social world that has been largely neglected.


Introduction
There are times when I need to ride in the subway at rush hour or sit in a crowded movie housethat's what I mean by a humanity bath. As cattle must have salt to lick, I sometimes crave physical contact. (Bellow, Ravelstein, 2000) With the coronavirus pandemics and the lockdowns, people left the streets and confined themselves in their homes. We had to be physically isolated from each other. We could no longer enjoy a cup of coffee in a café or study in a library. I did not think I would miss being in a library. I always thought of studying in the library as an individual action. However, with the lockdown, I missed being near others. Humans are gregarious creatures, and this desire to be near others is a manifestation of our gregarious nature. It is the desire to share presence with others, to connectbut not necessarily a strong form of connection. It can be as simple as the desire to study in a place where others are also studying. The presence of other people sometimes is like a fireplace. Too near would be too hot. Too far would be too cold. Near, but with some distance, we get the warmth of sociality. Consider this example: Saul is alone at home when he feels like drinking coffee. He goes to the kitchen to prepare his own coffee, but he changes his mind and decides to go to a café near his home. Saul does not call any friend to go with him, and neither does he intend to meet other people in the café. He is thinking of sitting alone in the café saloon, where most likely there will be other people, minding their own business.
There seems to be a difference between drinking coffee alone at home and drinking coffee in a café. And yet, drinking coffee in a café is not a joint action. It is an individual action but performed in a social environment. Most theories in the literature on social ontology focus on interactions and, as such, tend to focus on the rationality of cooperation or the normative aspects of the social world. Granted, Saul's drinking coffee in the café is not a joint action, so we should not expect a theory of joint action to explain that phenomenon. However, some authors in the area, such as Margaret Gilbert (2014Gilbert ( , 2003, consider the concepts in their theories to be the fundamental block that composes the social world. Even though some authors, like Michael Bratman (2014), do not aim to explain the whole social world, they analyze social interaction without first considering the possibility of actions that are not completely individual (like drinking coffee alone at home) and neither completely joint (like sharing a cup of coffee with a partner). There is a gap in the literature. From the philosophy of action, which is focused mainly on individual agency, the literature jumped to analyzing joint action. 1 But there is something in between. My focus in this paper will be on these very light forms of sociality, such as going to a café, cinema, or library. This jump in the literature, from individual to joint action, has led many authors to consider many social situations as 'mere aggregations'. 2 Bratman (2014) used his theory of planned agency to explain shared agency, and focused on what he called 'modest sociality' (Bratman, 2014, p. 8). Modest sociality is about a small group of people coordinating in an intentional way. Coordination that does not involve an intention to act together would be mere aggregation. It became a widespread consensus that modest sociality is the most basic form of sociality. A consensus that has been accepted without questioning. Even the current trend to analyze minimal forms of joint action fails to capture sociality outside of coordination. As such, the field has become blind to lighter forms of sociality. We have been overlooking the gregarious aspect of aggregations.
Providing a theory of gregariousness will only enrich the current theories on social ontology. There are currently no theories on this type of sociality. Gregariousness is about the warmth of the social world. When we face lockdowns and cannot go to the cinema, study in the library, nor drink coffee in a café, what we miss is not just a normative relation with others. Neither is this a desire to engage in strategic interactions. Not being alone should be considered the most basic level of sociality. As gregarious creatures, we desire to share our presence. I will present four characteristics of gregariousness. The first is proximity, which is a feature of the environment, not a mental state (Section 2.1). The second characteristic of 2 Many authors in the literature on joint action use the term 'mere aggregation' meaning the simple sum of individual actions or attitudes, which would not amount to a joint action or a collective attitude. This notion is widespread. The following is just a short list of occurrences: Searle (1990, p. 402); Hammond (2016Hammond ( , p. 2711; Gilbert (2018, p. 224);Jankovic (2014, p. 498);Schmid (2016, p. 59);Jansen (2014, p. 92) ;Peter French (1979); Alonso (2017, p. 34);Caporael (1996, p. 268) ;Hakli, Miller and Tuomela (2010, p. 294).
Sara Rachel Chant contrasts 'aggregate actions from a mere collection of actions' (Chant, 2017, p. 20). What Chant means by aggregate action is different from collective intentional action (as usually conceived in the main theories, such as by Bratman or Gilbert). However, her notion of aggregation is very different from the gregariousness that I am analyzing in this paper. I will discuss Chant's theory in Section 3.1.
Christian List and Philip Pettit (2011) use the term 'aggregation'. However, their theory is about aggregating individual decisions into a collective deliberation. It is a different phenomenon from sitting in a café nearby other people, where there is no collective decision-making taking place. 437 gregariousness is openness (Section 2.2). Agents need to have an affable attitude toward each other. The third characteristic is asymmetry (Section 2.3). Agents need not have the same attitude, e.g. one can desire to form further joint actions whilst another agent desires to stay by themselves near others. It excludes only purely aggressive attitudes. This leads to the fourth characteristic, privacy (Section 2.4). I will also briefly explore the possibility of gregariousness in virtual environments (Section 2.5).
In Section 3, I will provide a negative thesis. I will start addressing how the literature overlooks the gregarious aspect in what they refer to as 'mere aggregation'. The essence of my negative argument is that gregariousness cannot be understood as coordination. This means that the concept of strategic interaction cannot explain gregariousness (Section 3.1). Then, I will explain how the concepts of shared planning agency, joint commitment and second-person standpoint can only explain the normative aspects of the café, but not gregariousness (Section 3.2, 3.3 and 3.4 respectively). Next, I will address the concept of institutional facts, which is based on the concept of we-intentions (Section 3.5). The concept of we-intention, however, has inspired the concept of sense of us, which could be weakened to provide an account of gregariousness (Section 3.6).

Positive thesis
Gregariousness comes from the Latin word grex, which means 'flock' or 'herd'. Gregarious creatures tend to form grex. Humans are gregarious creatures. We tend to form groups and stay nearby others. David Hume once wrote that there is no bigger punishment than being completely alone (Hume, 1978, p. 363). The state of affairs of being near others is what I am calling 'gregarious'. As such, gregariousness depends both on mental states (i.e. the attitudes of agents) and on the features of the external world (i.e. the environment). Since it requires more than just mental states, it is distinct from the concept of collective intention. As I will argue later, most accounts of collective intention rely solely on mental states. However, there is no grex if there is no proximity among the agents. 3 This proximity is not just physical. It requires the agents 3 What I mean by the term grex is something much weaker than a group. Víctor Fernández Castro and Elisabeth Pacherie (2021) have recently argued that the need to belong is what makes commitments credible and forceful. Inspired by Roy Baumeister and Mark Learye (1995) and Harriet Over 438 to recognize other agents and also not to treat them as enemies. Hence, gregariousness also relies on mental states. Gregariousness can be defined by four main characteristics: proximity, openness, asymmetry, and privacy. Let me explain each in more detail.

Proximity
The main difference between drinking coffee alone at home and drinking in a café is that, in the café, you are not alone. In order not to be alone, there has to be some proximity. By proximity, I mean actual proximity in the world. Proximity is not a mental state. It is a feature of the environment.
Proximity can be a vague concept. People who live in Berlin are nearer the Eiffel Tower than people living in Tokyo. Even though it is easier for a person in Berlin to access the Eiffel Tower, it is highly probable that many people in Berlin have never even visited Paris. To say that something is near does not reveal much unless there is a reference point.
The proximity required for gregariousness is the one that impacts the agents' perceptual present space. If two people are in front of each other at a table, they are present to each other. Each perceives the other. Each perceives that they are sharing the environment with the other. In a way, we are all sharing the same planet, but the impact a person in Japan has on a person in Mexico is insignificant. It is not captured by the agents. It does not affect their present space.
It is hard to pinpoint the limits of the present space of an agent. One possibility would be to consider the agent's body as the centre, and the further away other elements in the environment are, the slower would be their impact on the agent (i.e. the less they are available to the agent to interact or perceive). 4 We can divide this distance (2016), they conceive the need to belong as the need to keep long-lasting relations with others. This motivates us to keep frequent interactions. What Castro and Pacherie are analyzing is stronger than gregariousness. The café, and the examples used by Saul Bellow, are not cases where there are long-lasting relations. 4 Such an account could be similar to the concept of oikeiosis (see Engberg-Pedersen, 1990). Oikeiosis is the ancient stoic idea that people perceive ownership over their own bodies. This creates a fixed point of reference, from which the person perceives concentric circles of relationship to her. The person then perceives what else she owns, which would be the 439 into levels. The first level would be the peripersonal space, which is the space immediately next to the body. The people around the person are at immediate access. This does not mean they are more relevant. Imagine a fugitive is in a café, surrounded by other people. She sees in the distance a police officer coming towards the café. The police officer has far more impact on her than the people around her, who are largely ignoring her. However, the police officer is far, so it is not as immediate as the people in the café. The proximity required for gregariousness, in this account, would be when it reaches the threshold to be perceived as immediate. 5 On this account, gregariousness is a state of affairs relative to the person. For example, imagine a very large café. Larger than a football stadium. Imagine a person sitting at one edge of the café. The people sitting on the other edge are not perceived as immediate by her. Her environment affords her gregariousness because she is surrounded by other people, but not because of the people at the other edge of the café.

Openness
In order for proximity to generate gregariousness, the agents need an open, affable, friendly attitude. 6 If an agent cannot conceive the other as an agent, then it is impossible for that person to conceive that she shares a space with others. On top of that, the agents should not have a purely aggressive attitude. This would not generate gregariousness but conflict.
This openness refers to being affable towards the presence of other people. It ranges all the way from wanting to interact with others to just tolerating their presence. Gregariousness does not preclude the family, the people you live with, your home. Oikos is ancient Greek for home. Then the person perceives her relationship to her city, which is another, bigger circle. And so on, until the person perceives the circle that encompasses all of humanity. 5 Immediate proximity could be understood as 'here', which could be understood in terms of peripersonal space (see de Vignemont, 2021). Proximity required for gregariousness would be an extension of the 'here'. 6 One of the first occurrences of the term 'sociality' in English is in David Hartley's book Observations on Man, from the 18th century. Talking about the pleasures related to sympathy, Hartley defines sociality as 'the Pleasure which we take in the mere Company and Conversation of others, particularly of our Friends and Acquaintance, and which is attended with mutual Affability, Complaisance, and Candour' (Hartley, 1749, p. 472). emergence of stronger forms of sociality. It is possible to be in a café, willing to engage in conversations with others. Even if you do end up engaging in a conversation with someone in the café, there is still a gregarious state of affairs between you and the others who are not taking part in the conversation. Openness also refers to just appreciating the presence of other people. Consider, for example, sidewalk cafés, where an individual can enjoy his cup of coffee while watching people passing by. 7

Asymmetry
The gregarious attitude is asymmetrical. It does not require reciprocity. The decision to go to the café instead of drinking coffee at home is motivated by the desire not to be alone. I do not need other people to desire the same. They can have their own reasons to go to the café. If other people are aggressive to me, then, in that case, it would block the gregarious attitude. As long as they are at least neutral, it does not hinder my gregarious attitude.
Asymmetry is the main difference between the gregarious attitude and the concept of collective intention, which I will discuss at length later in this paper in Section 3.2. For example, shared intention, according to Bratman (2014), requires all members to have the same intention. Gregariousness, on the other hand, is compatible with many different attitudes, as long as they are not purely aggressive. Very little is required from the agents in order to reach gregariousness, which makes it not cognitively demanding, and a good candidate for a minimal form of sociality.

Privacy
In the passage quoted at the beginning of this paper, Saul Bellow expresses his desire for a 'humanity bath', and the examples he 7 Both these notions of openness have, in some measure, been discussed by the famous urbanist Jane Jacobs (2016) and the sociologist Lyn Lofland (2017). Jacobs talks about how cities, at least the good ones, provide a multitude of random encounters. She discusses how sidewalks provide a place for brief non-intimate encounters and how important they are (Jacobs, 2016, pp. 33-34). Regarding Lofland, I will discuss her account in more detail later in Section 3.6. About sideway cafés, I also recommend Jan Oosterman (1992). 441 used do not usually involve further joint actions. When you take the underground, most likely, you do not interact with other people, no further than 'doing nothing together'. During the lockdown, I missed studying in the library. I missed being near others, studying in a place where other people are also studying. However, that specific desire was not to engage in conversations in the library. It was the desire to be in a social environment but to preserve a large degree of privacy. In the café example, I might not desire to talk to anyone else. Yet, if there was nobody else in the café, then it would not satisfy my desire. The same goes for the cinema. When the cinema is completely empty, then it is much like watching a movie alone at home, but with a better screen. A better screen is not what makes cinema what it is.
In the literature in social ontology, the analogy of 'glue' is commonly used (especially by Epstein, 2014). However, glue is not a good metaphor for gregariousness. Gregariousness is like gravity. It can pull to contact and be the initial force that turns into more stable forms of collective activity. But it can also be a force that makes people stay around, like satellites orbiting each other. Never in contact, but always close enough. Not isolated from each other, but in between others. Social, but not joint.

Gregariousness in Virtual Places
So far, I have considered physical places, such as cafés and libraries. However, nowadays, we spend a good amount of time on the Internet, such as on Twitter, YouTube and Zoom. One could argue that these platforms might provide the warmth of sociality, the feeling of not being alone. Consider, for example, the phenomenon of 'study with me'. These are students who live stream themselves studying for hours on YouTube, usually not showing their faces and not interacting with the viewers. Millions of people 'watch' these videos. They let it play while they also study. It provides them with a background noise of an environment of study. 8 These streamings can indeed provide some social warmth, but they are not gregarious state of affairs. This is because we are not sharing an environment. Consider the following example of a strange library. It is composed of a series of booths arranged in a circle, and in the middle of the circle, there is a desk. The surrounding booths are 8 Thanks to an anonymous referee for pressing me to develop more on this point. completely closed, except for a door to enter the booth and a two-way mirror on the front wall, which faces the desk in the centre. Two-way mirrors are transparent on one side but reflective on the other (they are very common in detective movies). The students in the surrounding booths cannot see each other, but they can see the student at the desk in the centre through the two-way mirror. The student at the desk cannot see anyone. This strange place looks more like a panopticon than a normal library. There is no shared environment. There is no proximity. Streamings of 'study with me' are more similar to that strange library than to normal libraries. Also, it is worth noting that although there are plenty of 'study with me' and 'code with me' streamings, there are no 'drink coffee with me', where the streamer drinks coffee without addressing the viewers and people at their homes also drink coffee without paying too much attention to the streaming. Usually, the streamings of '"activity" with me' are about motivating others to study, code or read. 9 I am not saying that these streamings cannot provide feelings of togetherness, nor that it is impossible to have gregariousness through the Internet. Lucy Osler (2020) has analyzed at length communal experiences online. There are several online forums, and they can form a group identity and become a community. Real-life communities also extend online, such as a group of friends using Facebook. However, in order to provide gregariousness, we need to actually share an environment. This environment can be virtual. Online games, specifically the massively multiplayer online games with a persistent world, are the best example of virtual gregariousness. In these games, each player has an avatar, and the players do not need to interact. Nevertheless, they share an environment, and it is very common for players to play these games without engaging in joint activities (the so-called 'solo players').
Osler also talks about how online games can provide a sense of togetherness (Osler, 2020, pp. 583-584). However, she focuses on the interaction, such as accomplishing missions together. Massive multiplayer online games typically involve interactions, but it is also typical for players to play alone. Some of the 'solo players' do not even like to 9 Maybe it is possible to conceive a group of socialities that do not involve joint action, of which gregariousness would be a sub-set. Another sub-set would be doing the same thing, but not together, as in the case of the "study with me". In this latter case, it is about doing the same thing. That situation cannot be explained as shared planning because it would not satisfy all of Bratman's conditions (such as common knowledge) and it does not involve joint commitments, but there is something social about it. 443 talk to other players. It is interesting that these people desire to play a multiplayer game, though they do not want to 'play together'. My point is that they have a desire for gregariousness. They want to play where others are playing, and maybe they enjoy seeing other avatars nearby, doing their own things. This experience is rather different from watching someone streaming themselves playing a single-player game and you playing as well at the same time (which would be akin to the 'study with me' phenomena). I am not saying such a situation would not involve any form of sociality, but only that they are different things. I do not have space to provide a more thorough analysis of virtual gregariousness, as it would require a more detailed analysis of what a virtual environment means and the distinction between a platform like YouTube and an online game with a persistent world. And it would also require a thorough analysis of gregariousness, whereas, in this paper, I am only providing a prospective account.

Negative thesis
As I have been arguing in this paper, there is a gap between individual action in isolation and collective action. This gap will not be filled by analyzing minimal forms of joint action, for one main reason. They are bound to the notion of interaction and coordination. Bratman (2014) is one of the most influential authors in this area. He coined the term modest sociality to refer to simple forms of group coordination, e.g. pushing a piano upstairs. He considers these as 'basic forms of sociality' (Bratman, 2014, p. 3). Anything less than that would be a 'mere aggregation'. This view is widespread, but it should be questioned because it overlooks any form of sociality that fails to satisfy the criteria for joint action. Even accounts of minimal joint action face the same problem. Most accounts of minimal joint action start their analyses by identifying essential aspects of robust cases of joint action, like pushing a piano upstairs, and then weakening them to apply to simpler forms of joint action (Ludwig, 2020;Paternotte, 2020;Saint-Germier et al., 2021). However, no matter how much we weaken these aspects, they will always be bound to the interaction among the agents. 10 The aim of the arguments in this section is not to refute these views but rather to show that these theories fail to fill the gap that I am concerned with in this paper. I will start by addressing the concept of strategic interaction. Then I will move on to address the concept of collective intentionand show how these concepts are inappropriate for explaining gregariousness.

Not strategic equilibrium
Some authors have presented theories of joint action based on the concept of strategic equilibrium. Sara Rachel Chant (2017) argues that we can better explain joint action if we understand it as aggregate action. What Chant understands by aggregation is rather different from what I mean by gregariousness.
Chant presents two examples that would normally be considered joint action, but theories of collective intention would fail to explain them: Lost in Paris and Spy-proof Factory (Chant, 2017, p. 20). The Lost in Paris example is about two friends who lose each other in Paris, and each decides to go to the Eiffel Tower in the hope of finding the other. This is a classic example used by Schelling to explain salience in game theory literature (see Schelling, 1963, andLewis, 1969). The Spy-proof Factory example is about a factory where each worker knows only about their own task in the assembly line, they have no connection to other workers, and they do not even know what they are producing. These examples are problematic for both Bratman's and Gilbert's theories, Chant argues, because they do not fulfil the common knowledge condition in these theories. While Chant is correct in this assessment, note, however, how the examples are cases of joint action stripped from any gregariousness. The Lost in Paris is an example of individual rational agents in a pure coordination dilemma. The Spy-proof Factory is an example of extreme alienation in the workplace. Surely, we could conceive it as a joint action. Yet, it would be the coldest form of joint action.
The Spy-proof Factory would be a purely rationalized form of coordinating in society. There is no sense of proximity and openness. The example of Lost in Paris is precisely an example of not sharing a space, and thus not being present to each other. Strategic equilibrium can provide an explanation of how isolated rational decisionmakers can find a way to reach each other, which would then be a state of being together. However, it does not explain what the state of being together is.

Not shared intention
If we use the concept of shared planning agency to explain the sociality in the café, the only thing that is shared would be the policy not to bother others. Being near others without any desire to do so is a common thing. 11 In these cases, Bratman's notion of shared planning agency might be enough to explain the social interaction. What his theory cannot explain is the positive social aspect of the intention to drink coffee in a café and not alone at home.
Bratman argues that collective intentions can be reduced to individual intentions, so he prefers to call it shared intention. The idea, in short, is that we can explain shared intention by using the same elements used to explain individual intention. Intentions, in contrast to desires or goals, have four norms of rationality: means-end coherence, agglomeration, consistency with other intentions and beliefs, and stability (Bratman, 2014, p. 15). Intentions can explain how agents can perform complex tasks that extend over time. In a similar way, shared intentions are intentions that extend socially. Each member needs to include a reference to each other's intentions and sub-plans regarding the joint activity. 12 This creates a web of intentions that allows members to coordinate. When two friends are walking together, it is not required for all of their intentions to be means-end coherent 11 This notion of forced socialization can be found on Martin Heidegger's notion of 'being-with' (mitsein in German). He thought that human existence is always among other people (Heidegger, 2010). We are always surrounded by othersit is not an option. However, this is a notion of 'sharing presence' as a necessity. Heidegger's work in Being and Time inspired Jean-Paul Sartre's philosophy. For Sartre (2003), there is also a mode of existence that is social, it is the being-for-others. However, what Sartre means by being-for-others is different from Heidegger's being-with. Sartre talks about how a person feels shame when she becomes aware that someone else can see her. As Sartre argues, the presence of another person (or the Other, as Sartre calls it) objectifies us. This is a notion of sociality as an obligation, as something we have to do (as well represented in his story No Exit, Sartre, 1947). Later in life, Sartre changed some of his views, and he started to consider that social relations can have positive reciprocity, where each strengthens the freedom of the other (see Critique of Dialectical Reason, 2006, and Notebooks for an Ethics, 1992). But even in that case, the focus is either on joint projects or on not repressing other people. In this sense, applying Sartre's theory to the café would be rather similar to Darwall's second-person or Gilbert's mutual recognition of co-presence, which I will discuss in sections 3.3 and 3.4. 12 The summary of Bratman's theory can be found in Bratman) 2014, pp. 85-86).
with one another, nor that they are consistent or agglomerate with one another. What matters are only the intentions regarding the joint activitythese should be coherent and consistent with one another in an intentional way.
Even though Bratman explicitly says that he has only provided sufficient conditions for shared intention (i.e. not necessary conditions; see Bratman, 2014, p. 36) and that he is focused on modest sociality (i.e. small groups of adults with stable membership, without hierarchy imbalance), the only thing his theory can explain is the normative aspect of being near other people, but not the four characteristics of gregariousness. Let me explain this by using an example from Velleman (2015).
When you walk completely alone, that action is individual. It does not involve any kind of reference to anyone else. When you walk in a crowded street, you might not share an intention with other people, but you need to be aware of the presence of other people. You do not care about what others are doing, but you do not want to bump into others. Bumping into others is not just morally wrong. It is also inefficient. Consider car traffic. People want to get back home, so they do not want traffic congestion. Therefore, at the very least, there is coordination not to interfere with one another. If so, then there is a shared intention between the people in a crowded street, namely, the shared intention of doing nothing together (Velleman, 2015, p. 108). 13 It not only guarantees an efficient flux in the crowd but also protects each person from engaging in a shared intention with others. If I start walking next to you, it might give you the wrong impression that I want to walk with you, but I do not. The same could be said about the café, where each person might desire to enjoy their coffee without engaging in conversation with others. 14 Here is how one could use the concept of shared planning agency to explain 'doing nothing together'. Each person in a café has sufficient reasons to believe that everyone else there has the intention to mind their own business and not bother other people. They can only mind their own business in peace if everyone does not bother others. It is not much of a stretch to suppose that this is common 13 Velleman's argument that there is a shared intention to do nothing together fits Bratman's account, specially Bratman's account of shared policies (Bratman, 2014, p. 139). I should acknowledge, however, that Velleman has some criticisms of Bratman's theory (see Velleman, 1997) 14 What Velleman is talking about is very similar to Erving Goffman's notion of civil inattention (Goffman, 1980, pp. 83-84). I will return to Goffman's (1980) work later when I discuss Stephen Darwall's (2011) account, in Section 3.4. 447 knowledge among everyone in the café. In this case, there is a shared plan which provides the basis for each person to pursue their own private intentions. This shared plan of 'not bothering others' will operate as a shared policy (Bratman, 2014, p. 136). According to Bratman, intentions need to be able to agglomerate (Bratman, 2014, p. 15). This means that a person cannot rationally uphold conflicting intentions. The moment we form an intention, it usually becomes a filter for other intentions. For example, when I form the intention to go on a diet, it does not specify what I should eat, but it rules out the intention to eat junk food. In the café, there is the shared intention 'not to bother others', which acts as a filter for other private intentions. Thus, I can drink my coffee as long as I am not bothering others.
Bratman has argued that joint action does not necessarily involve some normative element, like joint commitment (Bratman, 2014, pp. 118-120). Nonetheless, shared policies are pretty much social norms. If we use the concept of shared planning agency to explain 'doing nothing with others', all we can explain is the normative aspect of not bothering others. It overlooks gregariousness. The difference between walking absolutely alone and walking in a crowded street would be that, in the latter, there is a normative requirement not to bump into others or too near others. This would mean that when Saul Bellow said that he needed to be in a crowd, it would be a need for sharing a policy, which pretty much amounts to a need for feeling constrained by a norm. However, this is not what Bellow needs. He needs 'a humanity bath'. Maybe he does not need to actually touch other people, but he needs to be near them. He needs to walk in the busy street. When I miss drinking coffee in a café, it is not the case that I miss being constrained by a norm. What I miss is being near others without engaging in further shared intentions with them. There is more to not being alone than just sharing a policy.
I am not arguing that a café without such a shared policy would be able to function as a café. Norms are important. My argument is that the concept of shared planning agency cannot grasp the positive aspects of gregariousness. There can be norms without gregariousness. I have presented earlier the example of everyone having to stay at home during the lockdowns. Even if there was no official law dictating people to stay at home, there could be a shared policy not to go out unless for essential things. When we stayed at home, we were being constrained by a norm. We were participating in a shared policy. Yet, there was little gregariousness. The concept of shared policy cannot grasp the difference between coordinating due to a norm and being near others 'doing nothing'.

Not joint commitment
Gilbert (2014) explains joint action using the concept of joint commitment. 15 To create a joint commitment, all parties have to agree (even if just tacitly, without overt communication, see Gilbert, 2014, p. 29;and Tollefsen, 2015, pp. 44-45). This is because a joint commitment is nothing more than a decision. Any decision creates a commitment (a commitment of the will, in her words). When you agree to walk with your friend, you two form a decision (a pool of wills), which generates a commitmentin this case, a joint commitment. This means that even if you change your mind and form a divergent intention, you are still part of the joint commitment.
Inspired by Charles Taylor's (1985) notion of entre-nous, Gilbert analyzes what could be the most basic form of joint commitment: mutual recognition of co-presence. Gilbert uses the example of a library (Gilbert, 2014, p. 329). Suppose I am reading a newspaper, and you are reading your book. Due to the lack of empty seats, we are sharing the same table. We are sitting across from one another. As you sit, you look at me and give a smile. I notice that you are looking at me, and I look at you, answering with a smile. Gilbert argues that this brief interaction establishes a joint commitment to recognize each other's presence. That brief exchange of looks was enough to communicate to each other the message that 'I see you, and I will respect your presence'. 16 If, after that brief interaction, I spread my newspaper over your face, you will rebuke me for not respecting your space. What entitles your rebuke is that we formed a joint commitment to respect each other's presence. If that brief interaction had not happened, then I could claim that I did not see you, but from now on, I will respect your space. You can rebuke anyone who bothers you, but there is a difference between rebuking someone who does not perceive your presence (so it is accidental) and rebuking someone who has already perceived your presence (this amounts to a lack of respect).
For Gilbert, what binds the participants of a joint activity is the joint commitmentit is what makes an activity collective. She says that it is the atom of the social world (Gilbert, 2014, p. 18;. In the café example, there is a joint commitment to do nothing 15 'Persons X and Y collectively intend to perform action A (for short, to do A) if and only if they are jointly committed to intend as a body to do A' (Gilbert, 2014, p. 83). 16 Although Gilbert does not refer to Goffman's work, her notion of mutual recognition is very similar to Goffman's notion of civil inattention, which I will discuss in the next section (Section 3.4).

449
together, but, as I have argued, there is more. The theory of joint commitment cannot explain the non-normative aspects of gregariousness. Moreover, it reduces 'sharing presence' to a normative relation, thus overlooking the importance of affability and amenity involved in being in a social environment. Think of a cinema. Even though I cannot rebuke other people for leaving the cinema during the movie, I can demand other people to be quiet if they are too loud. After many months without being able to go to a cinema, I find myself missing the experience of watching movies in the cinema. Maybe what I miss is the big screen or the audio system. Or maybe what I miss is eavesdropping on other people's chats after the movie. I miss hearing other people laughing when there is a joke or screaming when there is a jump scare. Maybe I miss the smell of popcorn and all the little noises that compose the social experience of watching a movie in the cinema. But most certainly, I do not miss being in a position to demand other people to stay quiet. 17 There is more to sharing a space than only norms.
Sharing space can involve conflict and might not be an amicable experience. However, that side of the social world has been analyzed at length by so many theories. Joint commitment is just one of them that can explain how we are able to stay civilized. The problem is that we have been overlooking the amiable side of the social world, and we need new concepts to explain it.

Not second-person standpoint
Darwall's notion of second-person relation is similar to Gilbert's theory of joint commitment. The main difference in Gilbert's theory is that she does not consider joint commitments as being moral (Gilbert, 2014, p. 310). We can be jointly committed to performing something wrong.
When I drink coffee alone at home, there is only me. When I drink coffee in a café, there are other people. I am drinking my coffee, but I 17 Watching a movie in the cinema could be understood as a case of joint attention (see Campbell, 2005). I might desire to jointly watch the movie. Discussing joint attention is beyond the scope of this paper. It suffices to say that being in a café is not a case of joint attention. Joint attention might require sharing presence, but sharing a presence is not the same as joint attention. I am tasting my coffee, and this is not a joint experience between me and the other people in the café who are not drinking coffee. Each person is attending at something different, but we are sharing the space. 450 am also present to you. We are in the presence of each other. Darwall argues that 'presence is a second-personal notion' (Darwall, 2011, p. 15). A second-personal relation is a relation of mutual accountability. When I perceive that we are sharing presence, I perceive my own presence in the I-Thou perspective. I perceive myself as also being present to you. We are aware of each other's presence, so we are accountable to one another. Darwall distinguishes 'being with' from 'being around'. 'Being around' would be just being physically nearby each other. 'Being with' is when we engage in second-person relations. As Darwall argues, presence 'is a "forensic," that is, a normative, moral space of second-personal interaction' (Darwall, 2011, p. 15).
Both Darwall's notion of second-person standpoint and Gilbert's notion of mutual recognition of co-presence are very similar to Goffman's notion of civil inattention (Goffman, 1980). Goffman was an influential sociologist who provided a rich analysis of the micro-interactions we perform in society. For him, when we share an environment, we necessarily communicate (Goffman, 1980, p. 33). This can happen through verbal communication and it can happen with a mere exchange of looks. It also happens even without any interaction whatsoever. For example, the way we dress and behave sends a message about our attitude (Goffman, 1980, pp. 26-30). If I pick my nose in front of other people, I am sending a message that I do not care about their presence. The problem with Goffman's view is that, just like the other accounts, it ends up being about the normative aspects of sharing an environment (see Goffman, 1980, pp. 22-24 and 243). Consider a sidewalk café. There is a pleasure in seeing others walking by. Certainly, if we stare at someone, this will cross a line. But we are not simply recognizing the presence of others. We appreciate it. We enjoy it. We cannot reduce this activity to simply civil inattention. 18 In recent years, Darwall has been shifting his analysis to non-normative second-personal attitudes (Darwall, 2017;. He has already written about love and trust as second-personal attitudes of the heart (Darwall, 2017). He argues that love is second-personal because it addresses someone and it invites an answer, although it is not normative (Darwall, 2017, p. 47). Loving someone is an invitation, not a normative requirement, for that person to love you back 18 Goffman also uses the term 'mere aggregation' as opposed to a social group. He argues that situational properties (which are the rules that guide an individual when in public environments) 'transform the gathering itself from a mere aggregate of persons present into something akin to a little social group, a social reality in its own right' (Goffman, 1980, p. 196my emphasis). (Darwall, 2017, p. 47). This contrasts with recognizing each other's presence, which would be normative. Even though it is possible to exist non-normative second-personal attitudes, I do not think they can explain gregariousness. There are three potential problems. First, there would be the problem of identifying which kind of second-personal attitude would be going on in a café. Certainly, they would not be attitudes of love. Second, second-personal attitudes are always about addressing someone and being addressed. However, in a gregarious state of affairs, it is not clear who is being addressed. When I drink coffee in a café, my action of drinking my coffee is not addressed to anyone in there. The same goes for studying in a librarymy actions are not addressing anyone. One could argue, based on Goffman's account, that we are always communicating with others, such as by the way we dress. But in that case, the only attitudes that could be considered as addressing others would be the ones related to normativity, such as civil inattention. In addition, even if we accept Goffman's idea that the way we dress is already communicating our attitude to others, this kind of communication is not I-Thou (one-to-one) but communication to everyone who could be in the shared environment. Third, even in the cases where my action is about someone else, such as sitting in a café and watching other people walking, it is asymmetrical in a radical sense. The person sitting in the sidewalk café enjoys looking at the pedestrians, and this does not invite a likewise attitude from the pedestrians. Most likely, the pedestrians will either ignore him or just exert civil inattention on their part.

Not institutional fact
John Searle's theory, in short, is that joint actions involve we-intentions. Searle uses the example of two people playing violin in an orchestra to illustrate what collective intentionality is (Searle, 1995, p. 24). The example contrasts two violinists intentionally playing together (i.e. the orchestra) and two violinists who unintentionally find themselves playing in such a way that the final result is similar to the orchestra. Searle is contrasting collective intentionality with 'mere aggregation'. This is similar to the case of a crowded street used by Velleman. If we use Searle's theory, we will have to say each one's individual intention is not derived from a collective intention to walk together (i.e. no joint action), but there is the 'background' to respect each other (i.e. 'doing nothing to each other').
This 'background' is the context upon which the intention takes place. Searle uses the example of a café to explain his point. I ask for 452 a cup of coffee, and the waiter brings it (Searle, 1995, p. 3). The café involves a series of collective intentions, named we-intentions in his account. Searle uses the term social facts 'to refer to any fact involving collective intentionality' (Searle, 1995, p. 27). An orchestra playing a song is a social fact. Some social facts are institutions. 19 An institution is composed of rules of the form 'X counts as Y in C' (Searle, 1995, p. 43).
These institutional facts compose the background of most social activities. It provides a set of expectations and readiness for certain behaviour, e.g. people do not normally approach others in a café. To this extent, the notion of institutional facts can explain some difference between drinking coffee alone at home and going to a café. However, this background of institutional facts is, essentially, a set of norms and guides for behaviour. Moreover, it requires reciprocal attitudes (as it requires we-intentions), and it does not require proximity, whereas gregariousness admits asymmetrical attitudes and it requires proximity.
When Saul Bellow says that he misses being in a crowd, he misses the 'warmth' of the crowd. The violinist case used by Searle is a good example here. What Searle calls a 'mere aggregation' might, in effect, be gregarious. A person can desire to practice the violin near others because she wants to feel the warmth of the grex. She wants to play the violin, but she does not want to be alone. If it so happens that there are other violinists nearby her, it could be even better for her, in case she has a very open, affable attitude.
The concepts of shared planning agency, joint commitment and institutional facts are hugely famous in the literature in social ontology. Yet, none of them gives the proper importance to the basic human desire not to be alone, the desire to feel somehow connected or near others. This is because these concepts were created to explain coordination in joint action. There is, however, one potentially interesting aspect to the concept of we-intention. Institutional facts rely on weintentions. We-intentions are intentions of individuals, but they require a non-reductive notion of 'us'. Having a we-intention might involve a sense of us, a concept that might be able to explain gregariousness.

Not sense of us, but similar
The sociologist Lyn Lofland (2017) is, perhaps, the author that got closest to identifying gregariousness. Inspired by the works of 19 There can be collective intentions without institutions (i.e., they do not need an 'institutional' background nor do they give rise to institutions). For example, a pack of hyenas hunting a lion (Searle, 1995, p. 38). It is a social fact, since it is collective intention, but it is not an institutional fact. 453 Goffman (1980) and Jacobs (2016), Lofland provides an account of the public realm. 'Public realm' is what she calls urban areas where strangers co-exist, such as a street downtown. Her notion of public realm overlaps with what I call a gregarious state of affairs, but, ultimately, they are different. For example, public realm is strictly about the co-presence of strangers (Lofland, 2017, p.14), whereas gregariousness can happen between the co-presence of friends (e.g. friends studying in a library but each at a different table). Public realm refers to all sorts of social activities that happen between strangers in a city, so she does talk about classic forms of joint action (although without engaging with the philosophical literature on the topic). Gregariousness, on the other hand, is precisely about the sociality that exists even when there are no joint actions. Nevertheless, her work can certainly inspire accounts of gregariousness.
What is interesting about Lofland's account is that she puts emphasis not only on the normativity required to preserve peace and privacy in public, but also on the pleasures of the public realm (Lofland, 2017, p. 77-78). Of the pleasures she identifies, two are most relevant: people-watching and public solitude. People-watching is about the pleasure of seeing others living their lives (Lofland, 2017, pp. 90-92). This can mean not only watching people walking, but also eavesdropping on a couple having an argument, for example. She also talks about the pleasure of public solitude, which is the pleasure of being alone but 'surrounded by the hum of conversation' (Lofland, 2017, p. 89). In my account presented here, that kind of pleasure is related to privacy and openness, as I presented in Sections 2.1 and 2.2. When Lofland talks about that kind of pleasure, she mentions the term 'sense of oneness'. It is the joy of being part of a crowd. She mentions an example of people who go to a specific café in Israel on the Memorial Day, when they ring the sirens to remember past tragedies (Lofland, 2017, pp. 89-90). Being in that café when this happens brings a 'sense of oneness with the other inhabitants' (Lofland, 2017, p. 89). Lofland's mistake was not perceiving that she quickly shifted from analyzing a gregarious state of affairs (e.g. being alone in a crowded restaurant) to analyzing a case of interaction (e.g. rising for a minute of silence in a specific place and time when everyone is supposed to do the same). In addition, Lofland also did not explore in more detail what this sense of oneness means. It is odd that she puts the sense of oneness as part of the pleasure she classified as public solitude.
I believe this 'sense of oneness' can be understood as the 'sense of us', as analyzed by Dan Zahavi (2015). Watching a movie in a cinema 454 is, arguably, a shared experience. We are all watching the same screen. So, when someone says that they miss going to the cinema, they might mean that they miss having a we-experience of watching a movie. This is how Zahavi explains what a we-experience is: Emotional sharing requires a preservation of plurality and a certain self-other differentiation but if the difference between self and other is too salient, it will prevent any experience of togetherness. (…) You need to experience the others' perspectives on you, you need to be aware of them as being aware of you and to see yourself through their eyes, so that you can come to experience yourself in the same manner as you experience them. When that happens, you can become aware of yourself as one of them or, rather and more accurately, you can become aware of yourself as one of us. (Zahavi, 2015, pp. 94-95) Maybe the definition above is too strong to explain the sociality in the cinema or the café (not the Israeli café that Lofland talks about, but the café example I presented at the beginning of the paper). It would require everyone to actively engage in being aware of one another, whereas, in a cinema, you do not know very well how many other people there are, who they are, or even if they are aware of you. If I am alone in the cinema, I normally eavesdrop on other people's conversations about the movie. I do not want to talk to them, I do not want to take part in the conversation, so I try to be discrete. It is not that I am hiding my eavesdropping action. It is just that I do not want to cross the line of privacy. Returning to the café example, we might ask ourselves what experience is being shared. I am drinking my coffee. You are using the wi-fi. We are not sharing the experience of drinking coffee, and we are not sharing the experience of using your computer. What is shared is the space. We experience the presence of one another.
The 'sense of us' involves empathy, and emotional sharing, which is not present (or at least not strong enough) in the café example, or in the library. I do not need to see myself through the eyes of the other person in the café. It might even be a stretch to say that 'I can become aware of myself as one of us'. Imagine a group of friends who are meeting in the café. They are a group. I am also in the café, sharing the space, but I came alone and I am not part of that group. They are aware of my presence. Still, maybe they would rather reserve the term 'one of us' only for the group of friends.
If we focus on the normative relation between the people, there is a risk of understanding the sense of us as a second-person standpoint, 455 as mutual recognition of co-presence. However, we can focus on the non-forensic aspect of sharing presence. Maybe it is possible to have a lighter definition of 'sense of us' (of we-experience), one that could accommodate the cinema example. This weakened sense of us cannot be understood in terms of I-Thou, i.e. as a web of one-toone relations, as I have argued before (see Section 3.4). Instead of using a metaphor of eyes (as in the passage above: 'to see yourself through their eyes'my emphasis), the metaphor could be about ears and hearing. Eyes encounter each other, as in face-to-face interactions of civil inattention, whereas hearing is something more diffuse. So, it is not so much about being aware of each other individual in the crowd (the grex, the flock), but just being aware that you are in a crowd. So, instead of a 'sense of us', we can talk about a 'sense of gregariousness'. To experience a sense of gregariousness would involve a stronger sense of self-other differentiation, but not strong enough to 'prevent any experience of togetherness'. In short, my argument here is that as long as we do not focus on the forensic aspect of the 'sense of us', nor constrain it as an I-Thou relation, it is possible to deflate that concept to provide a phenomenological account of gregariousness.

Conclusion
In this paper, I provided a prospective positive thesis of gregariousness, identifying four aspects: (a) It is the desire not to be alonethis means proximity. (b) It is openit does not have to be 'doing nothing together'; it can be a willingness to engage in further joint activities or just the pleasure of being in a crowd. (c) It is asymmetricalgregariousness requires affable attitudes, in that it is incompatible with a purely aggressive attitude. However, it does not require the same attitude, e.g. some might be willing to engage in joint actions whilst others just want to be near, and others might just tolerate the situation. (d) It maintains privacy. Proximity and asymmetry are the main reasons why the concept of collective intention cannot explain gregariousness. I also briefly explored the possibility of virtual gregariousness, and my argument was that it is only possible if there is a shared virtual environment. In the negative part of the paper, I have explained how the concepts of strategic equilibrium, shared planning agency, joint commitment, second-person standpoint and we-intention fail to explain gregariousness. I explored how Zahavi's notion of 'sense of us' could be weakened to accommodate gregariousness. 456 I have not explored the negative aspects of gregariousness. Gregariousness requires proximity, and this can be bad in case there is a risk (such as a deadly virus that spreads through the air). Also, gregarious state of affairs involve privacy. I argued that gregariousness does not preclude the emergence of stronger forms of sociality. However, there are some gregarious places that involve too much privacy, which makes it difficult for people to approach one another (e.g. a library). The sociality present in gregariousness might not be sufficient for a person not to feel lonely. It is all too common to hear people complaining that, in big cities, there are no meaningful interactions. Mark Twain, way back in 1867, said that New York 'is a splendid deserta domed and steepled solitude, where the stranger is lonely in the midst of a million of his race' (Twain and Meltzer, 2002, p. 82). Humans need more than just not being alone. We need to make meaningful connections, form intimacy and be part of well-established groups. However, the negative experience of the lockdowns supports the idea that the so-called 'meaningless sociality' so common in big cities is actually quite important. Humans are complex and delicate creatures. We need a whole array of different forms of socialities, as well as being alone sometimes.
Before the pandemics, I do not remember looking forward to taking a bus. It was something I had to do, and I would always avoid sitting nearby other people. Even though I do not miss getting a bus every day, there is something I miss. The almost complete absence of sharing space with others reveals that sharing space is more than normative relations. There is a quality to it. It is this quality that makes me choose to drink coffee in a café and not at home.
Most theories in social ontology, especially on joint action, have focused on the rationality of coordinating or on normativity. They approach sociality in a rather cold way. They overlook the warmth of the social world. We share our presence with others not only because we have to. We also desire to. These theories focus on how humans have the cognitive tools to engage in social interactions, but they seem to forget the most basic: we do not want to be alone. This is true for many animals, not just humans. Like penguins and elephants, humans are gregarious creatures. It is at the foundational level of the social world. It is the ground upon which the social norms and conventions can build their castles. 20 20 I would like to thanks Thomas Smith for his support and invaluable comments throughout writing this paper.