FUTUWWA · THE CODE OF THE YOUNG
On preferring the other — and on the evidence that a coded group outperforms a talented set, which is weaker than it is usually reported.
Futuwwa was among the more practical things the Sufi world produced: a code of conduct for young men banded together, organized into guild-like associations, with initiation, obligations and an elder.
It was aimed at exactly the situation that groups of young people reliably produce — status contests, loyalty, the pressure to prove something — and it addressed it with a short set of demanded behaviours rather than with exhortation.
The reason it is useful here is not the content of the medieval code. It is the machinery: how a small group gets an actual code rather than a set of stated values nobody acts on.
At the centre of the code sits ithar: preferring the other to yourself, in small and concrete ways, as a habit rather than as a gesture.
It is demanding and it is behavioural, which is what distinguishes a code from a values statement. Not “we value each other” but the specific act of giving up the better thing, repeatedly, where it is noticed.
The associations transmitted it through belonging rather than instruction: a young person entered, was bound by obligations, and learned the conduct from people already doing it. Nobody was lectured into futuwwa.
The verse is descriptive of a specific group rather than a general command, and the tradition built a code on it.
What makes it usable is the second clause. Preferring others when you have plenty is generosity; doing it when you are short is the behaviour the code is actually about, and it is the one a group can observe and hold.
Futuwwa was not a book of maxims; it was an institution, and the institution is the interesting part.
There was an initiation, which meant membership was a threshold rather than a mood. There were specific obligations, few and nameable. There was an elder who did not enforce the code but embodied it and guarded the entrance. And membership could be lost.
That is a considerably better design than most attempts at group values. It is short, behavioural, entered deliberately, held by the group rather than by an authority, and it costs something to belong to.
The claim that a cohesive group outperforms a more talented set is usually stated with more confidence than the evidence supports, and the difficulty is causal direction.
Carron and colleagues’ meta-analysis found a moderate cohesion-performance relationship — and found the path from performance to cohesion at least as strong as the reverse.4 Winning makes groups cohesive at least as reliably as cohesion makes groups win, which is not what the motivational literature reports.
What is better supported is narrower and more useful: explicit, specific, few behavioural norms are more likely to be followed than broad value statements, and norms held by peers are followed more reliably than rules imposed by an authority.5
So the code machinery is the defensible part. The claim that it makes you win is not established, and the honest version is that it changes what the group is like to be in.
Two objections, and the first is structural rather than incidental.
The case against
Membership that costs something and can be lost is precisely what gives a code its force, and it is also what makes the group capable of closing around itself, treating outsiders as lesser, and punishing the member who raises a problem as disloyal.
The historical futuwwa associations were for young men, which is not incidental — the code was built for a particular kind of membership and its language assumes it. Lifting the machinery without noticing that is the same error the Aristotle collection had to state about friendship.
The guard is the Not-Two collection’s test and it belongs here: can the least senior member say the inconvenient thing? A code that has made that harder has not built cohesion, it has built a closed room.
Write five behaviours, not values. Observable enough that somebody could tell you whether you did them. “We support each other” is not one; “nobody leaves until the equipment is put away” is.
Have the group write them rather than the senior person. A code handed down is a rule, and rules are followed differently from norms.
Make one of them a version of preferring the other — something small and costly and visible.
And name explicitly who is expected to raise problems, with the expectation falling on somebody junior. That is what keeps the boundary from closing, and it has to be set up while things are going well.
The output is five behaviours owned by the group, and then a long time of actually holding them.
The useful part is the design: few, behavioural, written by the group, entered deliberately, held by peers, and costing something to belong to.
The claim that this wins is not established, and the direction of the evidence runs at least as much the other way. What it changes reliably is what the group is like to be inside.
And a code is a boundary. The test of whether yours is the good kind is whether the least senior person can still say the awkward thing.
Have the group write five behaviours, not values — observable enough that somebody could tell you whether you did them. Make one of them small, costly and visible.
Then name who is expected to raise problems, and make it somebody junior. Set that up while things are going well.
Arabic and Persian terms are transliterated simply. Where a saying is of contested authenticity the essay says so. Modern research is cited by author and publication.