BY 4.0 license Open Access Published by De Gruyter March 29, 2021

We in Me or Me in We? Collective Intentionality and Selfhood

Dan Zahavi ORCID logo


The article takes issue with the proposal that dominant accounts of collective intentionality suffer from an individualist bias and that one should instead reverse the order of explanation and give primacy to the we and the community. It discusses different versions of the community first view and argues that they fail because they operate with too simplistic a conception of what it means to be a self and misunderstand what it means to be (part of) a we. In presenting this argument, the article seeks to demonstrate that a thorough investigation of collective intentionality has to address the status and nature of the we, and that doing so will require an analysis of the relation between the we and the I, which in turn will call for a more explicit engagement with the question of selfhood than is customary in contemporary discussions of collective intentionality.

During the past few decades, collective intentionality has been explored in various disciplines including philosophy, social, cognitive and developmental psychology, economics, sociology, political theory, anthropology, ethology, and the social neurosciences. The most influential philosophical contributions by figures such as Searle, Bratman, Gilbert, and Tuomela have often engaged with issues in philosophy of action, and, in particular, addressed the question of how it is possible for individuals to collectively intend to do something, such as go for a walk or paint a house together. The available proposals then differ on whether or not they take collective intentions to be reducible to individual intentions, and on where exactly to locate the jointness in the collective intentions. By contrast, there has been considerably less philosophical focus on the question of how collective intentionality affects identity. But is that not a significant lacuna? After all, we are not simply individuals who on occasion act with others. We are also community members, shareholders in collective identities, and our being with others is often supported by various institutionalized, ritualized, and linguistically mediated norms and conventions. We identify with different groups, we share emotions, preferences, and values with other group members, and can come to feel, think and act as part of a we. But can we really understand what that amounts to, if we do not have a proper grasp of what it means to be a self? Does the first-person plural perspective presuppose, precede, preserve or abolish the first-person singular perspective? Is individual subjectivity something that necessarily requires a communal grounding or does a we presuppose a plurality of pre-existing selves? Whereas recent influential analyses of collective intentionality have primarily focused on joint action and paid far less attention to the link between collective intentionality and matters of identity, much of the early formative work on collective intentionality from the first decades of the 20th century recognized that a thorough account of collective intentionality would have to get clearer on the relation between the individual and the group.

Consider, for instance, some examples first offered by the social psychologist William McDougall (1920, p. 56–58) in an attempt to distinguish different types of social experiences and group phenomena, and then further discussed by the sociologist Alfred Vierkandt (1923, p. 364):

  • A group of men is on a road leading across a wilderness to a walled city. A sudden threat of danger sends them flying in panic towards the city gate. There is no co-operation, but we are not merely faced with a number of individual actions, since the impulse to flee in each of them is intensified by the presence and actions of the others.

  • A number of pilgrims wish to reach a walled city – each of them for their own private purpose. All of them know that robbers along the way may pose a danger to single travellers. They consequently join forces and travel together, and as a result their actions display a high degree of co-operation. But their co-operation is only a strategic means to an individual end. Each of them is only choosing the most efficient way to realize their own goal.

  • An army of crusaders approaches a walled city that needs to be captured. Each crusader identifies with the army as a whole and prizes its reputation and success as an end in itself. Each crusader experiences himself as part of a we, and in seeking to capture the city the army can be said to exercise a collective will.

The claim made by both McDougall and Vierkandt is that these three examples target different kinds of explananda, and that the analysis of the last example, which they consider to be a case of a genuine group experience, will require a reference to and elucidation of the first-person plural.

We can find an attempt at doing exactly that in Gerda Walther’s doctoral dissertation Ein Beitrag zur Ontologie der sozialen Gemeinschaften from 1921, which still counts as one of the most comprehensive phenomenological investigations of group experiences and communities.

As Walther initially points out, if a plurality of individuals is to come together as a we, the individuals in question must have something in common, there must be something they share. But it is not enough that they have the same kind of intentional state and are directed at the same kind of object. These conditions could be satisfied even in situations where the individuals had no awareness or knowledge of each other. It would also not be sufficient to simply add the requirement that the individuals should interact, since we could imagine a situation where individuals suspicious of one another are interacting in pursuit of the same goal (Walther 1923, p. 31).

What also needs to be in place, according to Walther, is the presence of a certain inner bond or connection (innere Verbundenheit), a feeling of togetherness (Gefühl der Zusammengehörigkeit), or reciprocal unification (Wechseleinigung) (Walther 1923, p. 33, 63). Only then will the experiences and actions be felt by the participating individuals, not as yours and/or mine, but as ours, as experiences or actions that we are undergoing or performing together (Walther 1923, p. 75), and only then will the experiences and actions be truly joint and shared. If we wish to understand the nature of collective intentionality, it is consequently not enough to merely look at the intentional object or the intentional state. Walther also insists, correctly I believe, that we need to look at the subject side and at the processes of identification and unification that affects self-experience and transforms the very sense of ownership (see León and Zahavi 2016).

1 The Primacy of the We

Can one at all find similar ideas in the contemporary debate? Whereas the mainstream view by now is that collective intentionality cannot be explained as a mere summation or aggregation of individual intentionality, there remains widespread disagreement about where to locate the collectivity. Three prominent proposals target the content, mode and subject side respectively (Schweikard and Schmid 2013). Whereas the first account argues that the collectivity is located in the content, i.e., in what the individuals believe, intend, feel or hope, and the second argues that the collectivity is located in the mode, i.e., in how the individuals believe, intend, feel or hope, the third account specifically targets the identity of the bearer of the collective intentionality, and argues that collective intentionality presupposes the existence of a plural subject, a we.

Let us take a closer at the latter proposal, which prima facie seems to resonate best with the proposals of McDougall, Vierkandt and Walther.

A leading contemporary proponent of the subject approach is Margaret Gilbert. For Gilbert, we cannot account for collective intentionality and the many kinds of group formations that we encounter in everyday life unless we explicitly address the nature of the we (Gilbert 2014, p.5). One can, however, talk about a we in different ways. It can be used in a distributive sense as a mere stand in for ‘all of us’, as in “we all play solitaire” or “we are all hungry” but the term can also be used in a collective sense and refer to more than a mere aggregate of persons, as in “we play tennis together” or “we want to marry”. In these latter cases, the notion involves a sense of unity, it links the involved individuals in a way where they jointly constitute what Gilbert calls a plural subject. More specifically, Gilbert’s key idea is that a plurality of individuals can come to be unified and act as one as a result of a joint commitment (Gilbert 2014, p. 63–65).

Despite Gilbert’s explicit defense of and reference to a plural subject, despite her claim that collective intentional states have a single ontological bearer, her individualist leanings remain quite visible, however (see also Schmid 2009, p. 31). After all, on her account, the main challenge is to explain how something like collective intentionality is possible, how it can grow out of individual intentionality. In short, what we need is an explanation of how a number of individuals can come together as a we, and not an explanation of how each of us eventually manages to secure some degree of separation and independence. But perhaps this is a mistake? Perhaps it reflects a failure to properly assess the relation between self and community? Why assume that collective intentionality is harder to explain than individual intentionality, why assume that the first-person singular is more self-explanatory than the first-person plural? Is such an assumption not simply the expression of a widespread individualist bias? Might it even, as Annette Baier has argued, demonstrate the extent to which many of us have been subjected to a Cartesian brainwash (Baier 1997, p. 18)?

On Baier’s account, it is simply wrong to assume that the analysis of singular-person actions must precede that of plural-person actions. Many standard types of action require at least two actors. Tangoing and marrying are obvious examples, but think also of agreeing, bargaining, buying, complaining, condoling, confessing, confiding, consenting, and convincing. Perhaps individual action is the exception rather than the rule; something we all have to learn by departing from common action (Baier 1997, p. 29). Perhaps it is our separateness and not our togetherness that is puzzling and in need of an explanation. In the end, Baier endorses a social constructivist take on action and reasoning and argues that both action and reason are essentially interpersonal, and only derivatively and parasitically solitary and individual (Baier 1997, p. 35).

But what about self-identity and experience? What about ascribing primacy to the group and the plural self? Although this might seem a natural next step, it is not a move that Baier pursues in her article, and it is a step that even ardent defenders of the irreducible character of collective intentionality have typically refrained from taking. McDougall, for instance, was unwilling to accept the existence of a “collective or super-individual consciousness, somehow comprising the consciousness of the individuals composing the group” (McDougall 1920, p. 19), and when Searle insisted that the very notion of a group mind is “at best mysterious and at worst incoherent” (Searle 1990, p. 404) he was speaking for many (see also Gilbert 2014, p. 9, 119). But perhaps there is a way of defending the primacy of the we that can avoid what some would take to be excessive metaphysical commitments, e.g., a commitment to the existence of some kind of primordial “hive mind”. Why not simply endorse a version of what Robert Wilson has dubbed the social manifestation thesis (Wilson 2005, p. 229), i.e., the thesis that certain psychological properties and capacities are manifested only in certain kinds of social circumstances. One could then more specifically argue that the I – the first-person singular perspective, the self – is communally grounded and enabled. Even if many might find it natural to ascribe the autonomous and individual self an absolutely central role in action, rationality, and morality, even if many might find the notion of such a self both fundamental and primordial, perhaps this “cult of the self”, as Marcel Mauss argued in a by now classical article from 1938, is in fact a rather peculiar and quite recent Western invention (Mauss 1985, pp. 3, 22). As several cultural psychologists have since argued, in a non-Western context, one can easily find a more relational, collectivist or “groupist” conception of self, one according to which the self is seen as “an integral part of the collective” and as nothing “without the collective” (Markus and Kityama 1994, p. 570). Indeed, as one translation of the Nguni Bantu term ubuntu has it: “I am because we are”.

This community first view can take different forms. Some would argue that we first experience ourselves as part of a family, a tribe or group, and automatically partake in its way of life, before we develop our own individuality and distinct perspective on the world. Others would defend the view that human self- or personhood presupposes the possession of the first-person concept. It requires the capacity to conceive of oneself as oneself and the linguistic ability to use the first-person pronoun to refer to oneself. On this account, being an I requires concept possession and language acquisition and ultimately membership of a linguistic community. A third option would be to argue that the community does not merely condition what we experience, but also that we have experiences. On this account, subjective experiences are social constructs. Despite their differences, all three accounts would claim that the standard way of addressing the topic of collective intentionality suffers from an individualist bias.

Even if this claim is correct, it doesn’t follow that the bias is necessarily unjustified. But let me try to unpack the proposals further, since this will give us a better idea of what it might mean to ascribe primacy to the we. Let me start with the first two accounts – I will return to the third option later.

1.1 The Normative Self

On a widespread view, the self is a kind of being that can only exist within a normative space. One can find, for instance, this line of argument in the communitarian criticism of liberalism. For communitarians, such as Sandel and MacIntyre, liberalism is premised on a commitment to an asocial individualism that fundamentally misunderstands the relation between the individual and the community (see Mulhall and Swift 1996). We are not social atoms that only subsequently form social relationships with others because we deem that to be to our individual advantage and conducive to the realization of our own pre-social goals. Indeed, to imagine that each of us develops our own preferences – be they culinary, religious or political – in splendid isolation, to imagine, as Hobbes did, that we each emerge from the Earth like mushrooms without any obligations to each other (Hobbes 1998, p. 102) – is nothing but a fantasy. My goals and preferences, what has significance and meaning to me, is largely shaped by the community of which I am part. It provides me with a background against which more individual choices about how to live can be made. But even more importantly, my very identity is not something ready-made, something fixed by nature that simply awaits discovery. Rather, it is by forging an identity that I become a self. Selfhood is an achievement rather than a given, more a matter of culture than nature. It is by living a life in accordance with certain normative guidelines that I develop my own point of view on matters, and thereby acquire a distinct individuality. Who I am is a question of what matters to me and what I care about. This is why knowing that I am, say, pro-life and pro-gun rather than pro-choice tells you something about who I am. If I change my interests, political views, religion etc., I change as well. It is consequently not simply my preferences and values that are influenced by my community. No, it is my very self-identity. To think that one can get to the core of human selfhood by abstracting away from the social context is consequently to fundamentally mischaracterize the relation between the I and the we. Rather than being an antecedently individuated self, rather than being merely contingently embedded in a community, my identity as an individual has a communal origin. As MacIntyre writes,

I am someone’s son or daughter, someone else’s cousin or uncle; I am a citizen of this or that city, a member of this or that guild or profession; I belong to this tribe, that clan, this nation. […] As such I inherit from the past of my family, my city, my tribe, my nation, a variety of debts, inheritances, rightful expectation and obligations. These constitute the given of life, my moral starting-point. This in part is what gives my life its moral particularity (MacIntyre 2007, p. 220).

Since my normative orientation is an essential part of who I am, I cannot be a self on my own, but only with others, as a participant in a process of social experience and exchange. I am individuated as a result of my membership in a communal tradition.

1.2 The First-person Concept

We find a slightly different defence of the claim that human selfhood is socially derived and belongs to the domain of social ontology in the work of Lynne Ruder Baker. Baker, who prefers to speak of the person rather than of the self, first argues that human persons necessarily possess a first-person perspective, and then insists on the need for distinguishing two stages of the first-person perspective, which she calls the “rudimentary” and “robust” stage respectively (Baker 2015, p.79). Baker equates the rudimentary stage with a non-conceptual capacity for intentional behaviour; a capacity, however, that requires both consciousness and intentionality. Such a rudimentary stage is not only possessed by human beings, but also by many non-human animals. Indeed, for Baker, all sentient beings possess a subjective point of view (Baker 2000, p. 60, 67). The robust stage, by contrast, is unique to human beings and is identified as the ability to conceive of oneself as oneself from the first-person (Baker 2015, p. 79). For Baker, it is the possession of this capacity to conceptualize the distinction between self and nonself that makes us into full-blown persons. It allows for self-consciousness, it provides us with an inner life, it permits us to take responsibility for the things we do, to care about the future, and to rationally assess our goals and values (Baker 2015, p. 80). How do we come to possess such a robust stage? According to Baker, it requires mastery of the first-person concept and the linguistic ability to use the first-person pronoun to refer to oneself (Baker 2000, p. 67–68). Since this involves language-acquisition, which in turn presupposes a linguistic community, Baker can ultimately conclude that human personhood is constitutively dependent upon membership of a community (Baker 2015, p. 84).

If we follow Baker in defining personhood in terms of a capacity for critical deliberation, the endorsement of values, moral commitments etc., the claim that the human person is a social entity that requires communal membership does not really appear as a very controversial claim. However, her own distinction between the rudimentary and the robust stage of the first-person perspective gives rise to an obvious question. What about the rudimentary stage of the first-person perspective, which involved possession of phenomenal consciousness and intentionality? Is that also a social property? Is that also socially constituted? On the one hand, Baker does claim that we only have an “inner life” as a result of being (linguistic) community members (Baker 2015, p. 80). On the other hand, Baker admits that she cannot really prove that the first-person perspective at such (rather than merely its robust stage) is constitutively dependent upon social communities and communal membership (Baker 2015, p. 84–85). Is there not a tension here between making an “inner life” dependent upon the robust stage of the first-person perspective, while at the same time admitting that there is phenomenal consciousness already at the rudimentary stage? And even more importantly, will the latter concession really allow for a robust rejection of the individualist bias?

There are many good reasons to reject an aggregative account of community, i.e., the view that the relation between a community and its members can be understood in analogy with the relation between a heap of sand and its composite grains. But the two proposals just considered are neither convincing rebuttals of the individualist bias nor convincing defences of the primacy of the we. One reason for this shortcoming is that both operate with too undifferentiated conceptions of what it means to be a self. To see why, we need to take a brief look at the ongoing debate about the nature of self.

2 Varieties of Self

There has been, and continues to be, much controversy about the nature, structure, and reality of the self. However, one idea that is being increasingly discussed in both philosophy and empirical science is the idea that the self is neither simple nor univocal, but better viewed as multifaceted. Whereas William James already differentiated the material, social and spiritual self (James 1890, p. 292), and Ulrich Neisser distinguished the ecological, interpersonal, extended, private, and conceptual self (Neisser 1993), Antonio Damasio and Stan Klein have more recently argued that evidence from neuropsychology and neuropathology points to the multidimensional nature of self (Damasio 1999, p. 16–17, 127; Klein 2010). In a target article published in Journal of Consciousness Studies in 1999, Galen Strawson summarized the ongoing controversy about the notion of self by listing more than 20 notions, including autobiographical self, narrative self, core self, dialogical self, embodied self, normative self, and neural self (Strawson 1999, p. 484). As should be clear from these few references, the discussion is both complex and somewhat confusing. While different authors operate with slightly different distinctions and slightly different labels, there is, however, growing consensus that it makes sense to distinguish, at the very least, a more primitive experientially grounded self from a more normatively enriched and extended self (Fuchs 2017; Gallagher 2000; Strawson 2009; Zahavi 2005, 2014).

For present purposes, the main systematic point is the following. Some dimensions of self are clearly social and first established in and through development and socialization. This would, for instance, include those aspects that are constituted by the values and norms we endorse. These dimensions can also be lost, for instance, in severe dementia. But there are other, arguably more fundamental dimensions that are present from very early on and which are linked to our embodiment and experiential life. Consider, for instance, the fact that we encounter the world from an embodied perspective. The objects I perceive are perceived as being to the right or left of me, or as being within reach or further away from me. According to some developmental psychologists, from early infancy children are able to discriminate their own body from the surrounding environment; they can perceive where they are, how they are moving, what they are doing, and whether a given action is their own or not (Neisser 1993, p. 4). Likewise, our experiential life is not merely characterized by its qualitative features but also by its subjective character. There is not simply something it is like – qualitatively speaking – to taste coffee, to feel a headache, or to enjoy a movie. We do not experience thirst, pain, pleasure, drowsiness, and happiness as free-floating anonymous events, but as self-concerning experiences. When having a headache, I am not faced with a two-step process in which I first detect the presence of an unpleasant experience, and then wonder whose experience it might be. Rather, experiences are necessarily like something for a subject, they necessarily involve a point of view, they come with perspectival ownership. Experience entails what-it-is-likeness and what-it-is-likeness entails what-it-is-like-for-me-ness (Zahavi 2014, 2020; Zahavi and Kriegel 2016). It might consequently be argued that a minimal form of selfhood is a built-in feature of experiential life, and that only, as Joseph Margolis has put it, “the utter elimination of experience could possibly vindicate the elimination of selves” (Margolis 1988, p. 41).

A possible response from the side of communitarians to this more minimalist proposal is that it utterly fails to target and capture the full richness of human selfhood. But one could accept this criticism and still insist that the minimalist concept captures something important. Indeed, why see the two outlined notions of self as competing alternatives that we have to choose between, rather than as complementary notions that each target different aspects or levels of selfhood?

One classical articulation of such a composite view can be found in the work of Husserl, who distinguishes between what he calls the pure ego and the personal ego.[1] We can understand the notion of pure ego to refer to the very subjectivity of experience (Husserl 1989, p. 103), i.e., to the first-personal character of consciousness. Importantly, this feature, this peculiar mineness (Meinheit) of our experiential life, is, for Husserl, not socially constituted. It doesn’t come about as a result of a confrontation or interaction with others, but denotes the fundamental and intrinsic individuation of consciousness – an individuation that according to Husserl also prohibits any fusion between streams of consciousness (Husserl 1973b, p. 335, 351).

The pure ego has an important role to play in Husserl’s account of consciousness, but as he also points out, even though our experiential life is inherently individuated, we should realize that it is a fairly formal and empty kind of individuation (Husserl 1973a, p. 23). This can be brought out by the following consideration: I can come to have the same kind of experiences, thoughts, beliefs and preferences as somebody else without becoming the other, just as somebody else can come to have the same type of experiences and beliefs as I have without thereby becoming me. Given that this is the case, it cannot be the specific content of experience that constitutes my being as a subject and distinguishes me from others. Rather, my most basic self-identity is the formal identity of my pure ego, which precisely lacks any “explicatable content” (Husserl 1989, p. 191). An obvious implication of this is that getting a better understanding of my pure ego will not really provide me with much information about who I am and how I differ from others – except in the most formal sense of the term. But I am not merely a pure ego. I also have character traits, abilities, dispositions, interests, habits and convictions, and since this is all something that the pure ego lacks, the latter should not “be confused with the Ego as the real person, with the real subject of the real human being” (Husserl 1989, p. 110).

What about the personal ego then? For Husserl, our identity as persons, our personal character and individuality, is constituted by our personal convictions, commitments and decisions and is under constant development (Husserl 1977, p. 164). It is by being committed and devoted to a certain set of central values and by leading a life in the light of specific norms, that I come to have a view and voice of my own, that I come to be a true individual in the robust sense of the term. But again, how does this happen? Here Husserl’s answer is very different, since he explicitly argues that we develop into a person and as a person in and through social relations with others (Husserl 1973a, p. 175). Many of the convictions I come to hold are convictions I appropriate from other community members through processes of communication. What they take to be valid acquires validity for me as well. In arguing for this view, Husserl explicitly emphasizes that my being as a person is not simply my own achievement, but the result of what Husserl would call my “communicative intertwinement” with others (Husserl 1973b, p. 603; see also Zahavi 2019).

Why this digression? The point that I would like to get across is the following. If we wish to account for genuine group experiences, we need to address the status and nature of we. Doing so will require reflections on the relation between we and I. But when discussing that relation, it is crucial to be clear about what notion of self one is operating with. That clarity (and a more general and explicit engagement with the question of selfhood) is often missing in contemporary discussions of collective intentionality.[2] That one’s notion of self makes a difference to the argument, should be obvious. It is one thing to argue that an “individual achieves selfhood at that point at which he first begins to act toward himself in more or less the same fashion in which he acts toward other people” (McCall and Simmons 1978, p. 52) and on the basis of such an understanding of selfhood then to argue that “the self emerges in social interaction within the context of a complex differentiated society” (Burke and Stets 2009, p. 10) as symbolic interactionists in the tradition from Mead have done. It is something completely different to define the self in terms of the first-personal or subjective character of phenomenal consciousness, i.e., to claim that experiences qua conscious are characterized by a distinct subjective or self-involving character, and to then argue that this minimal type of self is socially constituted and that it presupposes a communal grounding.

The central question, however, is what claim an ardent defender of the primacy of the we must commit to. As suggested by my discussion of Baker, I think the answer is straightforward. If one really wishes to extirpate the so-called individualist bias one must go all the way and argue for the radical claim that experiences are socially constructed. Wolfgang Prinz, who defends such a view, has claimed that human beings who were denied all social interaction (like the famous case of Kasper Hauser) would be like zombies, “completely self-less and thus without consciousness” (Prinz 2003, p. 526). Comparable views have been defended by Suzanne Zeedyk who writes that subjectivity is born from intersubjectivity, and that it is the latter that gives rise to capacities such as “self-awareness, representation, language, and even consciousness” (Zeedyk 2006, p. 326), and by Aikaterina Fotopoulou and Manos Tsakiris, who argue that the “phenomenal quality of conscious states is interpersonally constituted” (Fotopoulou and Tsakiris 2017, p. 7).

Let me emphasize how radical a position we are dealing with. The claim is not merely that consciousness emerges in a social context, in the sense that creatures endowed with consciousness de facto live together with others in a public world from the very start. The claim is much stronger. It is that phenomenal consciousness is constitutively dependent upon social interaction. Not merely when it comes to its specific content, but as regards its very being. One blatant weakness of claims to this effect is that they rarely spell out precisely how social interaction is supposed to give rise to experience. Some developmental psychologists have suggested that it is the caretaker that teaches the infant to attend to its own initially non-conscious affective states and that the latter only become experientially manifest as a result of being introspectively monitored by the infant (Gergely 2007). The problem with this account, however, is that it – like all other higher-order representational accounts of consciousness – fails to explain how a non-conscious mental state by being targeted by another (in this case, socially induced) non-conscious higher-order mental state can be transformed into a subjective experience.

I have discussed such a constructivist take on consciousness on previous occasions (Zahavi 1999, 2004, 2014, 2020) and will not repeat the criticism here, except in order to say that I don’t find it too surprising that some recent defenders of this approach, including Jay Garfield (Garfield 2019), have explicitly come out as proponents of illusionism. Most of us think and believe that we have experiences of pain, sorrow, happiness and orgasm, but for illusionists the fact of the matter is that those beliefs are all false. We have simply been taken in by a cognitive illusion. The view ultimately being defended is consequently that there “is no phenomenal consciousness” (Garfield 2016, p. 73).

3 The Plurality Requirement

Let us for the sake of the argument assume that the radical view of the social constructivists is true. Would we then be in a better position to understand and explain genuine group experiences than some of the existing proposals with their alleged individualist bias? This is far from obvious. If we consider illusionism, it simply denies the existence of the explanandum. Since there are no experiences, there are no group experiences either, so there is nothing enigmatic that needs to be explained. Illusionism consequently has wide-reaching implications. If you eliminate the first-person singular, you also lose the first-person plural. But more realist positions also face serious problems.

You can be a full-fledged member of certain groups, say, the group of people with blood type O, regardless of whether or not you know or care about it, just as outsiders might classify you as a member of these groups quite independently of your own view on the matter. Such externally enforced classifications are, however, not of much relevance if we wish to understand what it means to be part of a we. In contrast to various kinds of aggregate groups, a we requires an experiential anchoring. To be part of a we, is to be with and relate to other prospective members in a distinct way; one involving participation and identification. If you don’t self-identify as a member of the relevant group, you are not part of that we. To put it differently, you cannot be a member of a we without somehow affirming or endorsing that membership experientially. This is what makes the we a first-person plural. Saying this is not to say that the identification with and participation in a given group always happens deliberatively and voluntarily or that it cannot be based on shared objective features such as biological kinship. One might be born into and be brought up within a certain family and community, and such memberships might be quite beyond the domain of personal will and decision. What is important, however, is that the membership in question involves rather than bypasses the self-understanding and first-person perspective of the involved parties. Even in such cases, for the membership in question to count as a we-membership, it requires that you do experience yourself as one of us. Qua first-person plural, however, a we also requires plurality. Even if members of a we must be bound together in some fashion, even if the togetherness distinctive of a we requires some kind of integration, we should not be looking for an undifferentiated fusional unity (León, Szanto, and Zahavi 2019). Rather, within a we, differences must be preserved and experienced in order to make possible a genuine being-with-one-another. One might express this by saying that the interpersonal differences must be bridged rather than erased. Heterogeneity is an essential part of communal life. But if this is so, the attempt to derive the I from the we, the suggestion that the we precedes and enables individual differentiation – be it on the level of identity or on the level of experience – must be rejected as incoherent.

The plurality requirement has been recognized by many contributors to the debate on collective intentionality. Gilbert, for instance, has argued that a we is a plurality of persons (Gilbert 2014, p. 9) and that the constitution of a plural subject “requires a plurality of individual participants” (Gilbert 2014, p. 238). Similar sentiments can be found in more classical accounts as well. In a text from 1938, for instance, Martin Buber writes, “By We I mean a community of several independent persons, who have reached a self and self-responsibility” (Buber 2002, p. 208).

Contrary to what might have been expected, my preliminary conclusion is consequently that the individualist bias that can be found in most contemporary approaches to collective intentionality is well motivated.[3] Not only do I think that attempts to ground phenomenal consciousness in communal life and to derive the very subjectivity of experience from group-membership fail, but it should also be recognized that individual minds rather than being obstacles to are indeed preconditions for genuine we-phenomena. To some extent, this conclusion mirrors a position I have been defending for many years, the view namely that a strong commitment to the first-personal character of consciousness, i.e., to the inherent and essential individuation of experiential life, is a necessary requirement for any proper account of intersubjectivity (Zahavi 1999, 2001, 2014). To conceive of the difference between self and other as a founded and derived difference, say, as a difference that arises out of an undifferentiated anonymous life, obscures that which has to be clarified, namely, intersubjectivity understood as a relation between subjects. In a similar manner, the attempt to derive the individuality of minds from a pre-existing undifferentiated group will not get us what we want, namely a proper account of the first-person plural.

Let us consider three options:

  1. The we is prior to the I and the you.

  2. The I, the you, and the we are equiprimordial and co-dependent.

  3. The I and the you are prior to the we.

I have already rejected option 1. If we are to speak meaningfully of a we, of a first-person plural, we need to preserve plurality and differentiation. To conceive of the we as an undifferentiated oneness is to misunderstand the very notion. But what about option 2? I don’t find this convincing either. It is by no means evident that each and every self-experience and you-experience necessarily requires or presupposes a concomitant we-experience. I can be aware of myself (for instance, as an embodied agent) without being reflectively or pre-reflectively aware of myself as member of a we, just as I can be aware that you are my antagonist without that awareness necessarily giving rise to a shared we-perspective. These considerations suggest that option 3 is the correct one. But perhaps this proposal will be met by resistance and dismissed as too abstract. It could be argued that we in real life are together with others from the very start, that we de facto are all embedded in sociality from the very beginning. So why not just acknowledge that the collective we is prior to the individual I, or that they at the very least arise together?

But let us not move too fast. First of all, we should distinguish factual co-occurrence (which nobody is denying) from constitutive interdependence (which is a much stronger – and theoretically interesting – claim). The former is no evidence for the latter. To put it differently, an infant’s first experiences typically occur in the company of others, but that does not show that its experiences are enabled or constituted by social interaction. Secondly, one should not make the mistake of assuming that one can prove the primacy of the communal we, simply by pointing to the importance of sociality. Whereas different forms of we (a family, a soccer team, a religious community, etc.) are all quite particular social formations, sociality is a much wider umbrella term that also covers relations of hostility, indifference, instrumental interactions etc. To put it differently, the number of people with whom we have social relations is much larger than the number of people together with whom we constitute a we. Even if infants are ultra-social from birth onwards, this hardly shows them to be part of a we from the outset (Brinck, Reddy, and Zahavi 2017).

Denying that our self-identity can be reduced to or exhaustively explained by our group membership(s) is not to deny that this membership in many ways shapes who we are. Even if it turns out that a we requires some pre-existing (minimal) form of selfhood, it is far from obvious that genuine we-phenomena are compatible with just any account of self. It is, for instance, hard to see how overly solipsistic and disembodied accounts of mind and self would allow for genuine group-experiences. Selfhood is after all not only what allows us to mark our difference to others, it is also something that permits us to adopt and share a perspective with them. To put it differently, if it is acknowledged that we can come to share intentions, emotions, and even identities with others, this is bound to put pressure on various assumptions about the nature of selfhood and constrain the range of available options.

4 I, You and We

Nothing I have said so far should be taken to suggest that a focus on the we is wrongheaded. As was already suggested by McDougall, Vierkandt, and Walther, even if shared content might play a role, accounting for genuine group experiences does require us to address the first-person plural. But even though the we is an important explanandum that a thorough investigation of collective intentionality must address, it is not a fundamental explanans.

Arguing that a we requires a plurality of antecedently individuated selves, arguing that membership of a we will involve processes of group-identification that affect one’s self-experience is, however, only part of the story. Even if I cannot be a member of a we unless I identify with the group in question, my identification is only necessary and not sufficient for membership. Why is that? Because a we by necessity involves more than one member. And whether I count as one also depends on whether the others recognize me as such. An immigrant might feel completely assimilated and still be regarded as an outsider. To understand the nature of we, it is consequently not enough just to look at the relation between I and we. One also has to look closer at the relationship between the prospective members. To put it slightly differently, if we wish to understand what it means to share a belief, an intention, an emotional experience, or more generally, a perspective with others, we also need to look at how we relate to and understand each other in the first place.

It is beyond the scope of the present paper to discuss this theme in any detail, but let me make one suggestion that should illustrate how closely the different topics are interconnected.

In his article “Collective intentions and actions” Searle writes that you in order to have or act on collective intentions must suppose that “others are agents like yourself, that they have a similar awareness of you as an agent like themselves, and that these awarenesses coalesce into a sense of us as possible or actual collective agents” (Searle 1990, p. 414). Searle doesn’t elaborate on this, but the suggestion that collective intentionality requires some capacity for social cognition seems quite plausible. However, are all forms of interpersonal understanding equal to the task? Is it enough simply to be able to single out and relate to others as special kinds of objects (‘agents with intentions’)? Will two people who simultaneously adopt a third-person observer perspective on each other be able to adopt and maintain a joint we-perspective, or is something else needed? While recognizing that size matters – there are important differences between the kind of we whose members know each other in person and the kind of large-scale we whose members have never met, but who are nevertheless united via shared rituals, traditions and normative expectations – let me propose that second-person engagement is of crucial importance. To relate to and address another as a you (rather than as a he or she) is to relate to someone, an I, who in turn relates to me as a you. Second-person engagement is a subject–subject (you-me) relation where I am not only aware of and directed at the other and, at the same time, implicitly aware of myself in the accusative, as attended to or addressed by the other, but where the attitudes of mutual address establish a form of “communicative connectedness” (Eilan 2020, p. 8). It is a reciprocal exchange of address and response that affects and transforms the self-experience of the participating individuals (Zahavi 2015, 2019). Why might this be an important steppingstone towards the constitution of what is arguably the most basic type of we, the dyadic kind that exists between individuals who interact face-to-face in the here and now? A we is constituted by me and at least one co-subject. But for me to relate to another co-subject is precisely for me to relate to somebody who not only has a perspective of his or her own on the world, but on me as well.

The idea that the you is important for the we can not only be found in Buber – “Only men who are capable of truly saying Thou to one another can truly say We with one another” (Buber 2002, p. 208) –, but also in classical phenomenologists such as Husserl and Alfred Schutz. Schutz, for instance, claims that a we-relationship is established when two individuals engage in a reciprocal thou-orientation (Schutz and Luckmann 1973, p. 63).[4] And as he then continues:

In the we-relation our experiences are not only coordinated with one another, but are also reciprocally determined and related to one another. I experience myself through my consociate, and he experiences himself through me. The mirroring of self in the experience of the stranger (more exactly, in my grasp of the Other’s experience of me) is a constitutive element of the we-relation (Schutz and Luckmann 1973, p. 67).

To make this proposal compelling, far more details should be added, but it is striking that Schutz clearly considers a we-relation to involve a transformed self-experience and also believes an understanding of we to require a proper grasp of the intertwinement of the first-person singular, the second-person singular and the first-person plural perspective.

5 Conclusion

Contemporary discussions of collective intentionality have often focused on action and sought to explain how we can act together. How do we manage to paint a house, carry a piano, prepare a Hollandaise sauce, or go for a walk together? Although joint action is of obvious importance, an investigation of collective intentionality should not merely focus on the question of how individuals can share agentive intentions. We can also share emotions, beliefs, traditions, and identities, and a comprehensive analysis of collective intentionality must ultimately also clarify the nature of this we.

As I have argued in the preceding, one of the central questions that such a clarification must address concerns the relation between the individual and the community. Is the first-person plural independent of, and perhaps even prior to, individual subjectivity, or is it rather an achievement that has a first- (and second-) person singular perspective as its necessary precondition? I have presented arguments in favor of the latter view, but as I have also made clear much depends on the notion of self being employed. This is precisely why contemporary discussions of collective intentionality ought to engage much more explicitly with the question of selfhood than has been the case so far.

Corresponding author: Dan Zahavi, Center for Subjectivity Research, University of Copenhagen, Copenhagen, Denmark; and Philosophy, University of Oxford, Oxford, UK, E-mail:

Funding source: Carlsbergfondet

Award Identifier / Grant number: CF18-1107

Funding source: H2020 European Research Council

Award Identifier / Grant number: 832940


Thanks to Thomas Szanto, Felipe León, Mikko Salmela, Galen Strawson and in particular Alessandro Salice for comments to an earlier version of the article. I would also like to thank three anonymous reviewers for their helpful comments. This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 832940) and from the Carlsberg Foundation (Grant ID: CF18-1107).


Baier, A. 1997. “Doing Things with Others: The Mental Commons.” In Commonality and Particularity in Ethics, edited by L. Alanen, S. Heinämaa, and T. Wallgren, 15–44. London: Palgrave Macmillan. Search in Google Scholar

Baker, L. R. 2000. Persons and Bodies. Cambridge: Cambridge University Press. Search in Google Scholar

Baker, L. R. 2015. “Human Persons as Social Entities.” Journal of Social Ontology 1/1: 77–87, Search in Google Scholar

Brinck, I., V. Reddy, and D. Zahavi. 2017. “The Primacy of the ‘We’?” In Embodiment, Enaction, and Culture. Investigating the Constitution of the Shared World, edited by C. Durt, T. Fuchs, and C. Tewes, 131–47. Cambridge, MA: MIT Press. Search in Google Scholar

Buber, M. 2002. Between Man and Man, tr. R. G. Smith. London and New York: Routledge. Search in Google Scholar

Burke, P. J., and J. E. Stets. 2009. Identity Theory. Oxford: Oxford University Press. Search in Google Scholar

Damasio, A. 1999. The Feeling of what Happens. San Diego: Harcourt. Search in Google Scholar

Eilan, N. 2020. “Other I’s, Communication, and the Second Person.” Inquiry. Online first, Search in Google Scholar

Fotopoulou, A., and M. Tsakiris. 2017. “Mentalizing Homeostasis: The Social Origins of Interoceptive Inference.” Neuro-psychoanalysis 1/1: 3–28, Search in Google Scholar

Fuchs, T. 2017. “Self across Time: the Diachronic Unity of Bodily Existence.” Phenomenology and the Cognitive Sciences 16: 291–315, Search in Google Scholar

Gallagher, S. 2000. “Philosophical Conceptions of the Self: Implications for Cognitive Science.” Trends in Cognitive Sciences 4/1: 14–21, Search in Google Scholar

Garfield, J. L. 2016. “Illusionism and Givenness.” Journal of Consciousness Studies 23/11-12: 73–82. Search in Google Scholar

Garfield, J. L. 2019. “Second Persons and the Constitution of the First Person.” Humana.Mente - Journal of Philosophical Studies 36: 42–66. Search in Google Scholar

Gergely, G. 2007. “The Social Construction of the Subjective Self.” In Developmental Science and Psychoanalysis: Integration and Innovation, edited by L. C. Mayes, P. Fonagy, and M. Target, 45–82. London: Karnac. Search in Google Scholar

Gilbert, M. 2014. Joint Commitment: How We Make the Social World. Oxford: Oxford University Press. Search in Google Scholar

Heidegger, M. 2009. Logic as the Question Concerning the Essence of Language, tr. W. T. Gregory, and Y. Unna. Albany, NY: SUNY. Search in Google Scholar

Hobbes, T. 1998. On the Citizen. Cambridge: Cambridge University Press. Search in Google Scholar

Husserl, E. 1973a. “Zur Phänomenologie der Intersubjektivität.” In Texte aus dem Nachlass, Zweiter Teil: 1921–1928, edited by I. Kern. Husserliana, XIV. The Hague: Martinus Nijhoff. Search in Google Scholar

Husserl, E. 1973b. “Zur Phänomenologie der Intersubjektivität.” In Texte aus dem Nachlass, Dritter Teil: 1929–1935, edited by I. Kern. Husserliana, XV. The Hague: Martinus Nijhoff. Search in Google Scholar

Husserl, E. 1977. Phenomenological Psychology: Lectures, Summer Semester, 1925, trans. by J. Scanlon. The Hague: Martinus Nijhoff. Search in Google Scholar

Husserl, E. 1989. Ideas Pertaining to a Pure Phenomenology and to a Phenomenological Philosophy, Second Book: Studies in the Phenomenology of Constitution, trans. by R. Rojcewicz, and A. Schuwer. Dordrecht: Kluwer Academic Publishers. Search in Google Scholar

James, W. 1890. The Principles of Psychology, Vol. i–ii. London: Macmillan. Search in Google Scholar

Klein, S. B. 2010. “The Self: As a Construct in Psychology and Neuropsychological Evidence for its Multiplicity.” Wiley Interdisciplinary Reviews: Cognitive Science 1/2: 172–83, Search in Google Scholar

León, F., and D. Zahavi. 2016. “Phenomenology of Experiential Sharing: The Contribution of Schutz and Walther.” In The Phenomenological Approach to Social Reality: History, Concepts, Problems, edited by A. Salice, and H. B. Schmid, 219–34. Switzerland: Springer International Publishing. Search in Google Scholar

León, F., T. Szanto, and D. Zahavi. 2019. “Emotional Sharing and the Extended Mind.” Synthese 196/12: 4847–67, Search in Google Scholar

McCall, G. J., and J. L. Simmons. 1978. Identities and Interactions. New York: Free, Press. Search in Google Scholar

McDougall, W. 1920. The Group Mind. A Sketch of the Principles of Collective Psychology with Some Attempt to Apply Them to the Interpretation of National Life and Character. Cambridge: Cambridge University Press. Search in Google Scholar

MacIntyre, A. 2007. After Virtue: A Study in Moral Theory, 3rd ed. Notre Dame, IN: University of Notre Dame Press. Search in Google Scholar

Margolis, J. 1988. “Minds, Selves, and Persons.” Topoi 7 (1): 31–45. Search in Google Scholar

Markus, H. R., and S. Kityama. 1994. “A Collective Fear of the Collective: Implications for Selves and Theories of Selves.” Personality and Social Psychology Bulletin 20/5: 568–79, Search in Google Scholar

Mauss, M. 1985. “A Category of the Human Mind: the Notion of Person; the Notion of Self.” In The Category of the Person: Anthropology, Philosophy, History, edited by M. Carrithers, S. Collins, and S. Lukes, 1–25. Cambridge: Cambridge University Press. Search in Google Scholar

Mulhall, S., and A. Swift. 1996. Liberals and Communitarians, 2nd ed. Oxford: Blackwell. Search in Google Scholar

Neisser, U. 1993. “The Self Perceived.” In The Perceived Self: Ecological and Interpersonal Sources of Self-Knowledge, edited by U. Neisser, 3–21. New York: Cambridge University Press. Search in Google Scholar

Prinz, W. 2003. “Emerging Selves: Representational Foundations of Subjectivity.” Consciousness and Cognition 12/4: 515–28, Search in Google Scholar

Schmid, H. B. 2009. Plural Action. Essays in Philosophy and Social Science. Dordrecht: Springer. Search in Google Scholar

Schutz, A., and T. Luckmann. 1973. The Structures of the Life-World. London: Heineman. Search in Google Scholar

Schweikard, D. P., and H. B. Schmid. 2013. “Collective Intentionality.” In The Stanford Encyclopedia of Philosophy, Summer 2013 ed., edited by E. N. Zalta, Search in Google Scholar

Searle, J. R. 1990. “Collective Intentions and Actions.” In Intentions in Communication, edited by P. R. Cohen, J. Morgan, and M. E. Pollack, 401–15. Cambridge, MA: MIT Press. Search in Google Scholar

Strawson, G. 1999. “The Self and the SESMET.” In Models of the Self, edited by S. Gallagher, and J. Shear, 483–518. Thorverton: Imprint Academic. Search in Google Scholar

Strawson, G. 2009. Selves: An Essay in Revisionary Metaphysics. Oxford: Oxford, University Press. Search in Google Scholar

Vierkandt, A. 1923. Gesellschaftslehre. Hauptprobleme der philosophischen Soziologie. Stuttgart: Ferdinand Enke. Search in Google Scholar

Walther, G. 1923. “Zur Ontologie der sozialen Gemeinschaften.” Jahrbuch für Philosophie und Phänomenologische Forschung 6: 1–158. Search in Google Scholar

Wilson, R. A. 2005. “Collective Memory, Group Minds, and the Extended Mind Thesis.” Cognitive Processing 6/4: 227–36, Search in Google Scholar

Zahavi, D. 1999. Self-Awareness and Alterity: A Phenomenological Investigation. Evanston, Ill: Northwestern University Press. Search in Google Scholar

Zahavi, D. 2001. Husserl and Transcendental Intersubjectivity. Athens: Ohio University Press. Search in Google Scholar

Zahavi, D. 2004. “Back to Brentano?” Journal of Consciousness Studies 11/10–11: 66–87. Search in Google Scholar

Zahavi, D. 2005. Subjectivity and Selfhood: Investigating the First-person Perspective. Cambridge, Mass: MIT Press. Search in Google Scholar

Zahavi, D. 2014. Self and Other: Exploring Subjectivity, Empathy, and Shame. Oxford: Oxford University Press. Search in Google Scholar

Zahavi, D. 2015. “You, Me, and We: The Sharing of Emotional Experiences.” Journal of Consciousness Studies 22/1-2: 84–101. Search in Google Scholar

Zahavi, D. 2018. “Collective Intentionality and Plural Pre-reflective Self-Awareness.” Journal of Social Philosophy 49/1: 61–75, Search in Google Scholar

Zahavi, D. 2019. “Second-person Engagement, Self-alienation, Group-identification.” Topoi 38/1: 251–60, Search in Google Scholar

Zahavi, D. 2020. “Consciousness and Selfhood: Getting Clearer on For-Me-Ness and Mineness.” In The Oxford Handbook of the Philosophy of Consciousness, edited by U. Kriegel, 635–53. Oxford: Oxford University Press. Search in Google Scholar

Zahavi, D. 2021. “From No Ego to Pure Ego to Personal Ego.” In The Husserlian Mind, edited by H. Jacobs. London: Routledge. Search in Google Scholar

Zahavi, D., and U. Kriegel. 2016. “For-Me-Ness: What it is and What it is Not.” In Philosophy of Mind and Phenomenology: Conceptual and Empirical Approaches, edited by D. O. Dahlstrom, A. Elpidorou, and W. Hopp, 36–53. London: Routledge. Search in Google Scholar

Zeedyk, M. S. 2006. “From Intersubjectivity to Subjectivity: The Transformative Roles of Emotional Intimacy and Imitation.” Infant and Child Development 15/3: 321–44, Search in Google Scholar

Published Online: 2021-03-29

© 2021 Dan Zahavi, published by De Gruyter, Berlin/Boston

This work is licensed under the Creative Commons Attribution 4.0 International License.