People who cannot work together cannot succeed together.  

Mark Price



Features of cooperation

Separate interests become joint interest


Partner choice

Initiating collaboration

Working together

(“ought”: “we > me” → “others > me”)

See also: map of cooperation



Cooperation and the evolution of morality

Cooperation throws up various problems, for which morality, and a moral sense, are the solution.  See also: morality as cooperation

The human moral sense is a psychological matter, and our moral psychology has evolved to be this way through ancient selection pressures on our recent ancestor species.  

The psychological selection pressure on the entire human family tree was towards being able to learn, evolutionarily, how to cope with the demands of cooperation, in the form of collaborative hunting (from 500,000 years ago), foraging, child-rearing and providing that was obligatory because of the risky ancient human foraging niche of the African grasslands and open woodlands (from around 2 million years ago), where food is hard to obtain and process.  

The other great apes: chimpanzees, gorillas, orang utans and bonobos, do not have to cooperate (coordinate, communicate) in the same way to obtain their plentiful fruit or even to hunt for monkeys.  

The psychological mechanics of cooperation can be treated as a philosophical matter.  



Obligate collaboration and the evolved human sense of “ought”

The situation of obligate interdependence (a consequence of collaborating in order to survive) therefore provided the selection pressure that lead to the evolution of a number of “oughts”: things that humans today feel they should do, as part of their evolved moral psychology:  

  1. we “ought” to help others in need; we have evolved a desire to help others willingly (otherwise our partners or prospective partners might not be in good enough shape to collaborate with)
  2. we “ought” to stick to commitments (otherwise the project might fail, and we will let down our valued, respected and therefore deserving partner)
  3. we feel a sense of responsibility to our partners to uphold role ideals, and guilt when we fail.  
  4. we feel a sense of responsibility to all partners to share the spoils to our mutual satisfaction.  

Any failure to uphold an “ought” can result in protest from others, and damage to one’s cooperative identity; while virtuous behaviour builds our cooperative identity and can bring us praise.  





Personal interdependence and helping ("ought": "you > me")

Chimpanzees and bonobos help family, friends and coalitionary partners, but humans help collaborative partners (and prospective collaborative partners) too.  This is not necessarily based on calculations of returns, but just on a desire to help in certain contexts.  

On a small scale of cooperation, people are interdependent with known partners.  In this case, in the way that partner A needs partner B, it makes sense for partner A to personally help partner B in general, because partner A needs partner B to be in good shape.  This situation is experienced psychologically as a desire to help without necessarily receiving direct benefits in return.  See: stakeholder principle; unconditional love

It has been found experimentally in young children that their helping behaviour is not motivated by a desire for reciprocity (return benefits) but just to see the person in need helped, it does not matter by whom.  This means that the helping behaviour we see in humans is unlikely to have evolved in the context of tit-for-tat reciprocity (Tomasello 2016), and is likely to have evolved in a situation of interdependence and the need to keep people around us in good shape.  

Humans will “paternalistically” help in the sense that they will take care of someone's physical well-being before considering their other desires: for example, a cyclist who has fallen off their bike may be more concerned with their bike than themselves, but onlookers will wish to help the cyclist first.  This again suggests that the evolutionary origins of the extended helping behaviour of humans lies in interdependence.  

Altruism is one-way helping: X helps Y but Y does not necessarily help X in return.  

Collaboration is a form of two-way (or more) helping, where each partner helps the other to achieve a joint goal, including helping each other to fulfill their respective roles.  


Partner choice; reputation; mutual respect

People who cooperate do better than loners (Alvard 2013).  Humans need to collaborate in order to thrive and survive well, and consequently we face these two problems in social life: 1) finding good partners for cooperation: those who are skilled and diligent, and not lazy or dishonest, for example; 2) being chosen oneself for collaborative activities.  

We need therefore: 1) to know others’ track record as cooperators; 2) to have a good track record ourselves.  

Our image as a good or bad cooperator is called our reputation or cooperative identity.  As we evaluate others, we know that others are evaluating us.  This process of all evaluating all is internalised as a personal cooperative identity, defined as our opinion of ourselves as a cooperator.  Ultimately, because of the pressure to be chosen, there is a pressure to maintain our good cooperative identity.  A personal cooperative identity is roughly the same as a conscience.  

As a result of this situation of needing others to cooperate with, other people are potentially valuable to us, giving them bargaining power.  Conversely, we too are valuable to others and also have bargaining power.  This breeds mutual respect among potential partners.  

In response to the need to be seen to be a valuable partner, we engage in impression management, where we try actively to influence favourably how other people see us, and the best way to achieve this is actually to be a good partner, so we self-regulate in this direction.  


Initiating collaboration

We can initiate a collaborative activity either by:

  1. simply “falling” into it, or perhaps more usually, by
  2. making a joint commitment: explicitly communicating, between two or more people, an intention to collaborate to achieve a particular goal.  

To be trusted, a prospective partner has to have a good track record as a cooperator; i.e. a good cooperative identity or reputation.  One partner addresses the other respectfully, with an assumption that they are trustworthy and have a cooperative attitude.  

Explicitly communicating an intention to collaborate involves laying out in common ground knowledge 1) the goal; 2) all that is required to fulfill each role.  Each partner openly invites the other to make plans to do something, and to trust that they can be depended upon to persist in the goal until both are satisfied with the result.  Terminating the joint commitment requires another joint commitment in itself: one partner asks to terminate the commitment, and the other must agree.  

Any failure of the commitment on someone's part can result in damage to their personal or public cooperative identity.  


The content of the joint commitment is thus that each partner plays her collaborative role diligently and in the ideal way until both have benefited.  

Michael Tomasello – “A Natural History of Human Morality”


It is risky for partners to trust each other to stick to a collaborative activity faithfully to the end, without giving up, being tempted away or losing motivation.  What each needs is for the other to feel that they “ought” to stick to the commitment, that we owe it to one another.  Then, we can trust one other more deeply, in a more committed way, and the risk is reduced.  

There is a physical source of normativity (the joint goal) and a social source (other partners).  Normativity can be defined as “ought”-ness.  

The governing body in this situation, the social source of normativity (“ought”), is the joint agent “we” with which each partner identifies (“I am a member of the team”).

When the partners make the commitment, each relinquishes some control to the joint agent “we” (“we > me”): 1) for the sake of attaining the goal, and 2) because they care about their partners with whom they are interdependent (“you > me”).

This is internalised as a sense of responsibility to the others partners (“they’re my team”), backed up by the cooperative identities of all concerned.  

Partners therefore feel a responsibility to one another to achieve the goals of their role: the sub-goals of the overall goal.  



The joint agent “we”  

The joint agent “we” is the cooperating unit formed by individuals collaborating, coordinating and communicating towards a joint goal, cemented by a joint commitment.  

The individual members of the joint agent have joint attention, in that each is attending to the joint goal, together (i.e. each knows that the other is attending to it).

Each member has their own role to play and has an understanding of the other roles as well.  This means that in principle, the people playing the roles can be exchanged and the result can be the same.  

The same is true of perspectives: each player has their own perspective and has an understanding of the other perspectives too.  


Based on their capacity to coordinate actions and attention with others toward joint goals, humans came to understand that different individuals can have different perspectives on one and the same situation or entity.  ... We thus encounter once again the dual-level structure of simultaneous jointness [goal, attention, coordination] and individuality [roles, subgoals, perspectives].  Just as collaborative activities have the dual-level structure of joint goal and individual roles, joint attentional activities have the dual-level structure of joint attention and individual perspectives.  

Michael Tomasello – “A Natural History of Human Thinking”


It is possible that the members see the collaboration as a plan form, a “bird's eye view” of all the roles and perspectives within it, with their own just one among many.

Communication is used to skilfully coordinate the different perspectives and roles towards the joint goal.  The knowledge that is communicated then becomes part of the common ground knowledge of all concerned, and can be used by all to further the work towards the goal.  

These characters: “I”, “you” and “we” are the essential players in the morality of fairness.  “I” relinquish some control to “we”, which is internalised as responsibility towards “you”, with whom I feel mutual respect and deservingness, together with an impartial sense of self-other equivalence.  


Role ideals  or standards

A collaborative role ideal is the ideal way that a role should be performed in a particular activity, in order to ensure success.  They are the goals of each role: sub-goals of the overall goal.  There are general ideals that apply to all roles, such as helpfulness, diligence, honesty, faithfulness to the task, and fair sharing of the proceeds.  There is instrumental pressure towards achieving success and against failing.  

A role ideal – the virtuous performance of a particular role – is not just the goal of the individual player, but is a sub-goal of the joint agent “we”.  “We” is made up of individuals coordinating separate roles together to collaborate to achieve the joint goal.  

If an individual fails in their performance of a role, it negatively affects everyone involved in the project.  This common-ground knowledge adds a social pressure to perform properly, since we value our partners and what they think of us, and we also need to maintain our cooperative identities.  

A role ideal is “agent-independent” in that it applies impartially to anyone performing that role.  

Role ideals have all the hallmarks of being the ancestors of today's shared social norms.

General and specific role ideals

There are standards or ideals which apply to specific roles, and general ideals which apply to all collaborative roles alike.  These general role ideals include helping partners where needed; treating them with respect; sharing fairly; the general role ideal of upholding specific role ideals; seeing the collaboration through to the end until all have received their reward, without being distracted by something more interesting (sticking to the commitment); truth telling in communication, etc.  Therefore, cooperative moral principles (e.g. helping, fairness, upholding norms, honesty) can be seen as examples of general role ideals.   


Self-other equivalence (“you = me”)  

The existence of separate roles, each with its own necessary role ideals, leads to the fact that among partners, roles can in principle be reversed, and as long as the role ideals are upheld, the result will be the same.  Therefore the role ideals are “partner independent”, and must be upheld by any person who fills that role.  

The implication of this is recognition of a kind of self-other equivalence among collaborative partners: that within the roles, people can be interchangeable.  

Within a collaboration, each partner will necessarily swap perspectives with the other in order to monitor what they are doing, and also, in order to monitor what they themselves are doing from the other partner's perspective.  

In turn, this recognition of self-other equivalence, and perspective-swapping, has a number of implications for morality:  

Mutual deservingness, mutual respect, and impartial equality of status, make fairness possible (see Perfect Compassion).  

If somebody treats us badly we can say to them, “how would you like it if I did that to you?”  

If we see somebody in need, our empathic concern makes us trade their place imaginatively with a valued person X: ourselves or another loved one who may be similar to them in some way.  Then we say, “I wouldn’t like it if that happened to X”.  

We sometimes say to someone, “if I were in your position, I might have done the same thing.”  


Partner control and joint self-governance  

In partner control, we attempt to turn a failing or disrespectful partner into a good one, through “respectful protest”, or through helping and guidance, or a combination of the two.  (See also: unconditional love.)

There is positive pressure to treat the other partners well – the respect and deservingness we feel towards them.  There is also negative pressure in the form of the threat of sanctioning.  

Each partner relinquishes some of her personal control to the joint agent “we”, and in doing so, grants authority to the other partners to sanction her if she is wayward in carrying out her duties.  Through role reversal and self-other equivalence, she will also judge herself as deserving of their sanctioning under these circumstances, and if she is cooperative, will attempt to correct her behaviour in response.  

Each partner’s performance is therefore continually monitored and evaluated by the other partners, and by themselves, and each governs or regulates themselves in accordance with this evaluation.  

It is in the common ground knowledge of all partners that each is expecting to be treated with respect and as equals.  

If partner A feels that he has been treated unfairly, unjustly or disrespectfully by partner B, he can make an “respectful protest” towards partner B, informing her of his resentment but assuming that she is a cooperative person who wants to maintain her cooperative identity.  If partner B is still behaving poorly after this, then partner A always has an option to change partners (partner choice), and partner B will run the risk of damaging her own cooperative identity.  

This illustrates that the construction of a cooperative identity, and the process of morality itself, are interpersonal: both social and personal.  


Fair sharing and mutual deservingness

Sharing fairly means to be treated as an equal and given an equal share (equality) or an equal return on investment (proportionality).  We feel that other collaborative partners are deserving because


Partners to a joint commitment do not just prefer that we share equally; they feel that we owe it to one another to share equally.  

Michael Tomasello – “A Natural History of Human Morality”


Free riders have not done a share of the work towards our shared goal, are not part of our joint agent “we”, and have not helped us.  According to this, they do not deserve any of the rewards.  


Responsibility and guilt

I identify with our joint agent “we”, and relinquish some control to it (“we > me”), internalised as a responsibility to others to uphold the general and specific role ideals.  

If I fail in this regard then I will punish myself with feelings of guilt, and may elicit protest from others.  

Therefore, there is both a positive internal normative pressure (in the form of a sense of responsibility) and negative internal normative pressure (in the form of guilt) exerted by individual collaborative partners upon themselves.  






Alvard, Michael S. in Baumard, N; André, J; and Sperber, D (2013) – “A Mutualistic Approach to Morality: The Evolution of Fairness by Partner Choice”.  Behavioral and Brain Sciences, 36(1), p. 80

Tomasello, Michael – “A Natural History of Human Morality”: Harvard 2016

Tomasello, Michael – “A Natural History of Human Thinking”: Harvard 2014