Mining Super Cooperators
Mining Super Cooperators
This post is a coathanger for the book, so here is my takeaway. Onion dolls are us.
Martin Nowak with and Roger Highfield. Super Cooperators: Altruism, Evolution and Mathematics (or, Why We Need Each Other to Succeed). Melbourne: Text Publishing, 2011. 9781921758294
Picked this up sometime in 2020… —three years on my to-read shelf.
Even at a decade old, for me it is an excellent introduction or refresher on game theory and evolution, amidst some even more excellent stories about people around Martin Nowak’s mathematical life. I didn’t know about Kurt Gödel’s paranoia quite like that Freeman Dyson story on page 178.
I’ll have to follow-up on Corina Tarnita, or https://ctarnita.scholar.princeton.edu/
This post is a coathanger for the book, so here is my takeaway after reading it yesterday.
Mechanics of Cooperation
from page 270 (I am using substack’s “poetry block” to do this formatting :) )
① Repetition -- life goes on ② Reputation – Expected iterations or repetitions from ① (③ as terrain or season) ③ Spatial selection (this is the selection with a geography, not the selecting of a comfy spot)(landscape -- terrain as socially used/made/lived) ④ Multilevel selection (new term which includes group selection up to a point)(I might call it phenotypic norming or the fashions of identy). Have to do more reading. ⑤ Kin selection (at this point I realised these 5 points are not in any order)(forming)
① to ⑤ is actually an unordered list, the numbers are just identifiers as to how they are covered in the book, and so do not provide a taphonomy, nor a structure, so, then, let’s see, can we build it?
A (selfing/worlding)-centric way to do this with(-in) these co-evolving bits of crazy onion dolls that teach your grandmother to suck eggs:
ⓐ ⑤ Kin selection (Wittgenstein’s the “The world is given me”) You are no ‘me’ without others of your kind to be pronoun-ically distinct from. (Family as proto-group). ⓑ ④ Multilevel selection (Band-life where migration between groups is important)(no selection otherwise)(Human bands contain non-related individuals) Population survive, very indirect selection and is a process easily distracted by local events or perturbations. The non-relatedness ruins kin selection, which may not work on its own terms for eusocial insects anyways (says E.O. Wilson)). ⓒ ② Reputation – Expected iterations or repetitions (reciprocity/favour bank/insurance in band life-- narcissism policing) from ① Repetition ⓓ ③ Spatial selection (keep your frenemies close, & your unknown unknowns far far away)(Cities versus desert islands). ⓔ ① Repetition (eventually if you are lucky, and/or in the right place and time, you have a birthday)
Other pages
Page 110 “If, instead, you develop a model that is truly game-centered there is no need to use the concept of inclusive fitness at all.” —(inclusive fitness = group selection)(inclusive fitness == abstract-word abstractword = blahblah)
Reading page 94 —my notes — Because, group selection is at best a black box of taphonmically hidden processes, that we often get lost in just-so stories that tell us nothing. Group selection model may well work but never exclusively, their Achilles' heel is not their impossibility but the sea of options they swim in, options which may be simpler and do the job anyway. The question is not are they possible, but where are they probable, this is a “geographic” question (one of the answers might be the world) i.e. ③ Spatial selection related.
Other odds and sods
Multilevel selection is more interesting because it takes the possibilities of groups, as layers, like onion dolls, and tries to delineate their complexity and thus what emergence they may support/accrue. However, each doll inside a doll is treated as a POV and not as part of a (selfing/worlding) doing (living). As such a model like this may not extend into reality, only the mediating world, which is why group selection is so problematic.
The work Corina Tarnita has done on the multiple sets of groups we each live in (clubs, workplaces, sports, neighbourhoods, hobbies) is very interesting here. If you think of group selection we think of tribes or races, but those are adolescent versions of the multiple groups we live in. We police teenagers for this reason, it would be better not to have teenagers plonked together in prison like high schools I guess. Over-intensifying any cohort togetherness is a bad thing.
I work in a museum and the worst behaved groups of visitors are “groups” that at all the same in some way: teenagers, retireees, hens parties, drunk sailors. The worst is multiplied not the best (different groups membership, same structure same behaviour).
At some point in reading this I checked out Peter Singer’s “The Expanding Circle” and found a mention of a review that regards Singer’s work as a hybrid of Sociobiology by E. O. Wilson (one of Nowak’s cooperators on the kin selection rethinking) and John Rawls (who gets short shrift over at Equality without compromise : Liberal philosophy has clipped the wings of the egalitarian ideal. We should return to the bolder ideals of Iris Murdoch.)
But Singer’s Kantian big bang, is less interesting to me than that the shift from Kin selection to multilevel selection by Homo sp. (minus group selection) of the Band of unrelated individuals, where migration between groups is important to game-theory-ing policing defectors into more manageable, or at least survivable, parasites. I.E. we should be talking about band selection (by the band and between bands that band better than others) and not group selection in any case, and this neatly (i.e. messily but interestingly) overlaps/co-evolves with Corina Tarnita solution of Nowak’s sigma but for sets or clubs, not just Nowak & Antal’s Cooperators’ dilemma, but for multilevelled multi-setted (meta-) populations, after all once we get to band of unrelated individuals we already have multiple sets (or clubs) because band members (kith) come from different families (kin), so this is the norm for us later Homo, this is the first world where we are already given us.
We have always been here.
Our ancient many worlds expand and collide and absorb so there is only one world. Emperors confuse this with their empire, the one one oneness with one one leader one one people with one one belief. Such empires are an illness, but always with us, much like parasitic narcissists at the band level.
In terms of my learning about game theory, the main addition, in terms of practical worldbuilding is that punishment of bad behaviour is not rewarded, tit-for-tat is, but not tit-for-tat+something-extra (enjoying hurting people is a red flag).
