Wednesday, 7 March 2012

Tim Tyler: Nowak and Highfield, SuperCooperators (review)


Hi! I'm Tim Tyler and this is a brief review of this book:

SuperCooperators - Altruism, Evolution, and Why We Need Each Other to Succeed by Martin Nowak and Roger Highfield.

Martin Nowak is a Harvard biologist. His book is about why humans cooperate as much as they do - which has been a classic puzzle to evolutionary biologists for decades.

Nowak's number one answer seems to be "game theory". Nowak has worked on this topic with others in the field - including William Hamilton and Ed Wilson. In combination with Karl Sigmund, he came up with some of the ideas behind what he calls "indirect reciprocity".

The book has a chatty style that tries to weave the stories of those involved into the science. That might make the book easier to read for some, but it also makes getting at the scientific content harder. I found this aspect of the book pretty irritating. The author also focusses a lot on their own work and does not pay sufficient heed to the work of others.

Despite the author's reported enthusiasm for mathematical models, the book is almost entirely devoid of mathematics. I don't think I found a single equation - though there were some "sigmas" lying around.

Now on to some of my more serious technical issues:

The book starts out with chapters covering the basic mechanisms that the author claims produce cooperation. These are: direct reciprocity, indirect reciprocity, spatialisation, group selection and kin selection.

Spatialisation relies on reciprocity and kin selection to operate - and so barely deserves its own category, in my opinion.

Indirect reciprocity is a bit of a dustbin category - that mixes together some very different kinds of effects. I think it is better to separate these out - into virtue signalling - which is responsible for reputations and other ways of signalling niceness and overgeneralisation.

The list fails to mention the possibility that cooperative behaviour may have been adaptive in the past - when humans lived in close-knit tribes where kin selection and reciprocity may have played a bigger role. That's a pretty important scenario, in my opinion.

The list also fails to mention manipulation. Manipulation is a key mechanism promoting cooperation. If an agent takes a hit to help another agent, they may simply be being manipulated. They could be being manipulated by the other party, one of the other party's friends or relatives, or by one of their own enemies.

The list also fails to mention the symbiosis theory of cooperation. Parasites and symbiotes may manipulate their hosts into friendly interactions - since they depend on interactions between their hosts to spread.

Lastly, the list also fails to make any mention of the role of culture in forging human cooperation. The author has some awareness of this issue - since he has a chapter on the importance of language. Also he writes:

I do not restrict the use of the term “natural selection” to genes alone. Depending on whether we talk about cells, animals, or people, reproduction can be genetic or cultural. In the former case, successful individuals leave more offspring and pass more genes on to future generations. In the latter, successful ideas, fashions, and strategies spread by imitation and learning: a fad is born.

However, the treatment of culture in the book is cursory. Nowak views culture as facilitating reciprocal relationships via gossip and helping to distribute reputations. This limited perspective is unfortunate - since cultural transmission explains why modern humans are more cooperative than beasts - and the effects of culture aren't just down to reciprocity. In fact, cultural transmission is really what brings the symbiosis theory of cooperation to life. Memes manipulate their hosts into coming into close contact with each other - in order to facilitate their own reproduction. There are also cultural versions of kin selection and group selection to consider - in which these phenomena act on related memes and groups of memes respectively.

So, Nowak's list of mechanisms of cooperation misses out a lot that is significant. This book needs complete rewrite to account for the forces that actually result in human cooperation.

Nowak became infamous recently - in a bust up over kin selection. He wrote an article with E. O. Wilson about kin selection and group selection, in which kin selection got bashed. The article was called "The evolution of eusociality" and it was published in Nature. It was poorly written and was widely ridiculed. However, the group selection controvery and the role of E. O. Wilson in it is turning out to be an interesting chapter in the history of biology - and Wilson is publishing his own book on the topic later this year. One of the reasons I decided to read SuperCooperators is to see what Nowak had to say for himself on the topic in his book.

My council for any others thinking of doing the same is: don't bother. The section in the book bashing kin selection is pretty dreadful. Nowak doesn't say anything particularly interesting or insightful in the chapter that isn't in the "eusociality" paper. Rather he reinforces the impression that he doesn't know what he is talking about. He belittles Price's equation as being "tautological" - which seems to be a particularly stupid criticism to me. The chapter on group selection isn't much better. Group selection is an interesting topic - but I don't recommend learning about it from Nowak. He mostly just talks about the group selection models he personally built - without giving enough information for anyone to actually reproduce them. That is not worth very much, in my opinion.

In the final chapter, Nowak proposes that "natural cooperation" be added to Darwin's "mutation" and "natural selection" principles. I think that this suggestion doesn't make sense. Instead: production, elimination, mutation, symbiosis, synergy, self-organisation, learning and engineering are a much better list of fundamental evolutionary principles.

In the final chapter, Nowak explains why a scientific understanding cooperation is important for building cooperative enterprises - and that mastering cooperation is an important key for humanity when facing challenges in the modern world.

That all seems to be correct - but we need much better science that what Nowak has to offer in order to properly understand why humans cooperate.

There are a few interesting bits in this book, but probably its most salient feature is the way that it misses out so much that is of vital importance to the subject it is discussing.


No comments:

Post a Comment