Tuesday, 25 December 2012

The war on violent memes

Cultural epidemiology has found some new converts recently - in the area of combating violence. Here's cureviolence.org:

Treating violence as an infectious epidemic is effective

Three main strategies are used in reversing infectious epidemic processes. These are:

  • detecting and interrupting potential infectious events;
  • determining who are most likely to cause another infectious event and reducing their likelihood of developing disease and subsequently transmitting; and
  • changing the underlying social and behavioral norms, or environmental conditions, that directly relate to this infection.
These methods have resulted in reductions in shootings and killings of 16% to 34%.

The Cure Violence method is designed around these principles. This method begins with epidemiological analysis of the clusters involved and transmission dynamics, and uses several new categories of disease control workers – including violence interrupters, outreach behavior change agents, and community coordinators – to interrupt transmission to stop the spread and to change norms around the use of violence. Workers are trained as disease control workers, similar to tuberculosis workers or those looking for first cases of bird flu or SARS.

- http://cureviolence.org/what-we-do/

Another recent article on the topic is: Can Murder Be Tracked Like An Infectious Disease? by Shankar Vedantam.

Older articles on the topic include: Violence may be a 'socially infectious disease', Gun Violence Is Social Disease, Public Health Experts Say and Is It Time to Treat Violence Like a Contagious Disease?.

Meme inoculations, therapy and engineering have potential for treating other disorders besides violence. Drug abuse, road rage and many other mental health disorders could likely be treated as infectious diseases. However, first academics need to understand the idea of cultural epidemiology before they can properly investigate hypotheses based on it.

Friday, 21 December 2012

Humans vs Darwinism

One argument against memetics is that evolutionary theorists should not go in for controversy - and should instead close ranks to concentrate on their real foe: theism. I don't really approve of this argument - on the grounds that theism is has been scientifically dead for over a century, and there's no point is kicking a dead horse.

However, it is certainly true that some are still struggling with the first wave of the Darwinian revolution, let alone stages 2 and 3. A case in point is the recent works by philosophers:

Mind and Cosmos: Why the Materialist Neo-Darwinian Conception of Nature Is Almost Certainly False by Thomas Nagel and What Darwin Got Wrong by Jerry Fodor and Massimo Piattelli-Palmarini.

I'm inclined to blame such material on Sturgeon's Law. It is an illustration of human irrational resistance to Darwinism, though. Memetics brings Darwinism into human culture. The evolving brain puts Darwinism into the human mind. Some humans can't stomach this.

End of the world memes

Paranoia about the end of the world is an ancient phenomenon. Everyone wants to warn others about risks - and the end of the world acts as a risk superstimulus. Doomsday cults have capitalized on this effect throughout recorded history to spread their message, recruit new members and reallocate their resources. The result is large numbers of humans running around obsessed with the apocalypse.

In a curiously parallel phenomena, academic institutes have appeared which cater to those who fear the end of the world. The Cambridge Project for Existential Risk Oxford's existential-risk.org and Oxford Martin Programme on the Impacts of Future Technology. No doubt these will add credibility to the end of the world soothsayers and generally escalate the level of paranoia on the topic. There's also a high-tech cult - whose mission is to save the world from the coming apocalypse.

Whether the rise of the end of the world memes is desirable is debatable. Humans are naturally prone to paranoia due to the association between the ancestral environment and dangerous predators. Now we live in a comparatively safe environment, but out brains are still wired up as though we might die tomorrow. Humans are more jumpy and twitchy than makes sense, and feeding their natural paranoia seems unlikely to help them - instead causing them to waste their time and resources on insignificant risks.The last thing they need is large organizations feeding them risk superstimulii.

Indeed to combat such things, the US government now has a page for teaching parents how to reassure their kids in the face of the wave of apocalyptic memes they currently face.

The academics studying the topic suggest that many of the genuine risks facing humanity are caused by technological development. Technological development has in fact made the world a much safer place to live in for individuals. However, it should also be admitted that it has resulted in large nuclear stockpiles that could potentially cause large scale destruction of our planetary ecosystem. No such disaster has ever actually happened - so it is rather challenging to assess its probability.

The potential for technology to destroy the world seems likely to grow. No doubt there will be a parallel evolution of safeguards - to ensure that no such event never arises. However, I expect a corresponding escalation of end of the world memes - as we progress towards the next likely suspect: broad-spectrum superhuman machine intelligence.

The best memes of 2012

A roundup of lists from other sites:

The 'Buzzfeed' article says: "2012 will forever be known as the year the meme went mega-mainstream" - yay!

Saturday, 15 December 2012

Tim Tyler: Wilson, Darwin's Cathedral (review)


Hi. I'm Tim Tyler, and this is a review of this book:

Darwin's Cathedral: Evolution, Religion, and the Nature of Society by David Sloan Wilson

This is a relatively early book by a scientist looking at religion. Religion is a messy subject, which only a few scientists have attempted to deal with. Wilson's thesis is that religion is functional, and that its associated benefits accrue to groups of humans. He compares religious communities to beehives - on the grounds that their members cooperate with each other much as members of a bee hive do.

What are we to make of David's thesis? Bees in hives cooperate because they are extremely close relatives - since they are all daughters of the same queen. However most humans in religious communities are not anywhere near as closely related. So, in terms of DNA genes, the relationship between religious communities and bee hives is pretty far fetched. However, religious communities also share their memes, and a reasonable fraction of their memes is shared between community members. In particular the memes associated with their religion are often present in the form of near-identical copies in different members of the same religious community. Cultural relatedness does not necessarily lead towards altruism between the hosts involved, but it can do so. Much depends on the nature of memes in question and the strength of the host's memetic immune system. Memes don't have a free ride in manipulating the behaviour of their host - since they must compete with the other memes in the host and the host's own DNA genes. However with large coadapted meme complexes many memes can gang up together in an attempt to influence their host by force of numbers. That's exactly what religious memes do - and they evidently do have considerable success in influencing their host's behaviour. So, a comparison with bees may have something to it - though memetic relatedness between humans from the same religious group is probably not as high as genetic relatedness between bees within hives.

Another of the ideas David advocates is that religion is "functional" - by which he seems to mean adaptive to humans or groups of humans. He contrasts this position with "religion as a byproduct" hypotheses, economic theories involving religion as a form of transaction with alleged supernatural agents and the idea of religion as selfish memes. I think most consider religion to frequently be adaptive to its hosts. Religious people typically have more kids than secularists, often quite a lot more. One of the insights into the subject from cultural evolution is that when talking about the adaptive function of some aspect of religion, the DNA genes of the hosts are not the only possible beneficiary - religious traditions may be treated as cultural symbionts which have adaptations that benefit themselves. Wilson acknowledges the possible viability of such hypotheses, but categorises them in such a way that they compete with his own preferred explanation. He categorises adaptive theories of religion into those that invoke benefits to individuals, those that invoke benefits to groups, and those that treat religion as a cultural parasite that often evolves at the expense of individuals and groups. However, real religions vary considerably in the extent to which the interests of their memes is aligned with the the interests of the DNA genes of their hosts. Those religions which are transmitted primarily vertically down the generations can be expected to have evolved to have interests aligned with those of their hosts. Cultural and organic evolution pulling in the same direction explains the cases where religious groups typically have many children. By contrast, evangelical religions depend less on vertical transmission with respect to their hosts, and spread virally even between unrelated hosts. Such religions can be expected to be less in tune with the interests of their host's DNA genes, and more inclined towards redirecting host reproductive resources into meme propagation via evangelism. They will tend to be nastier religions.

David's correctly identifies kin selection at work - though he classifies it as group selection. Since group selection and kin selection are now widely thought to be equivalent, this is a valid perspective. However he doesn't really identify it as a cultural phenomenon. Indeed he seems to identify cultural evolution with the idea of "demonic memes" that act as parasites on humans - and then largely ignores it. Instead he proposes human groups as the beneficiaries of selection on religions. This seems like a muddled way of looking at the situation to me. Instead, the humans genes are weakly kin-selected, the religious memes are strongly kin selected - and the genes and the memes coevolve in a symbiosis. The interests of the memes and genes are somewhat aligned - largely due to the component of vertical transmission of religious beliefs. I felt that David's treatment of the topic muddled together cultural and organic evolution.

It is possible to ask whether religion is adaptive without distinguishing between cultural and organic evolution. I compare this approach to asking whether smallpox is adaptive. Through much of human history, smallpox helped groups of humans with smallpox to obliterate other tribes of humans who lacked it. Evidently smallpox is an adaptive trait at the group level. While partly accurate, this analysis is unorthodox - and misses out much of interest about the relationship between the smallpox virus and its human hosts. David's explanation of religion is like this. He just says it is adaptive at the group level - without teasing apart the relationship between the cultural and organic components of the system involved.

David does display some understanding of cultural evolution in this book. He invokes Calvin and Plotkin's idea of "Darwin Machines", uses it to explain how the brain evolves in a Darwinian fashion and then goes on to explain that human culture evolves. The section near the start of the book about cultural evolution is quite reasonable - as far as it goes.

Memetics isn't the only rival theory which I felt David treated unsympathetically. He also contrasts his approach with the idea of religion as a by-product. While functional explanations and "by-product" explanations can be seen as being opposed, it is pretty evident that the various "by-product" theories of religion have a lot going for them. The "Hyperactive Agent Detection Device" idea, is correct, for example. "By-product" hypotheses explain quite a few aspects of religion. Also, some of the traits which religion is thought to be a "by-product" of are themselves adaptive traits - so "by-product" hardly means the same as "non-adaptive". I think we should accept many of the "by-product" hypotheses concerning religion - without necessarily granting them everything.

It would be nice to have a scientific understanding of religion, not least so we can build new and better religions that draw from the best parts of their historical practices while missing out their toxic elements. However to do that we need to understand which bits of religion are desirable and which are not. Some things are obvious: yoga and meditation are good while hellfire and the oppression of women are not. However with other practices, things are not so clear. Just saying that religions are adaptive doesn't really help to identify which are the useful practices.

At the end of the book, David explains that a grant from the Templeton foundation helped to finance the book. The Templeton foundation is famous for paying scientists to say nice things about religion. I expect this funding source will turn off some readers.

Studying religion seems like a dirty job for a scientist - but someone has to do it. David Sloane Wilson seems to be OK with the topic. However back in 2002, he seemed to be rather hampered by his preference for explanations based on group selection and his reluctance to conceptually separate out cultural and organic evolution. Also, alas, this book isn't terribly readable. I found the long section analysing Calvinism in the middle to be especially tedious. I recommend that those interested in David's work should read Evolution for Everyone first.


Sunday, 9 December 2012

Tim Tyler: Whitfield, People Will Talk (review)


Hi. I'm Tim Tyler, and this is a review of this book:

People Will Talk: The Surprising Science of Reputation by John Whitfield

This is a science book all about the topic of reputation. Reputations are an important reason why humans cooperate. They lubricate reciprocal interactions by allowing creatures to consider a long history of an actor's behaviour towards others before dealing with them. Reputations are largely a product of sophisticated communication mechanisms and so represent something which humans have that most other animals lack.

I've been interested in the science of reputations for a while. In 2009, I made a video titled "universal karma" - which proposed we make more use of reputation systems, to better keep organisations in check, and for many other purposes.

This book is a popular science book covering what scientists know about reputation. The book is brilliant. It's well written, about a very important topic, and covers a good balance of subject areas. The author is evidently very smart, which always helps. Most of the book is devoted to explaining the science, but there are occasional sections about how to use the discoveries to improve the world.

The main message of the sections about how to change the world is that anonymity results in bad behaviour, while traceable identities and surveillance tend to make people behave well.

The book met or exceeded my expectations in practically every area. However, there were a few things I would have like to seen included that were omitted. I was expecting coverage of religious folklore oriented towards reputations. For example, the Hindu concept of Karma, or the idea that as you reap so you will sow. I also would have liked to see a bit of a historical perspective, showing the rise of reputation systems over time. The book does have a chapter about the internet era, and it's a pretty good one. I still wanted more though. Technology improves memory storage and facilitates surveillance - it has a powerful effect on the effectiveness of reputation systems.

The part of the book I was most irritated by was the section at the very end about global warming. The author is trying to find ways to apply reputations to big global problems, but I rate global warming as an awful choice of problem - it is a bad cause which already attracts far too much attention - without people trying to add further weight to it. Fortunately this section was short.

Overall, this book is pretty sweet. There aren't many science books on reputations and this is an excellent one - I recommend reading it.


Interview with the author about the book here.

Tuesday, 4 December 2012

Tim Tyler: Zarrella, Zarrella's Hierarchy of Contagiousness (review)


Hi. I'm Tim Tyler, and this is a review of this book:

Zarrella's Hierarchy of Contagiousness: The Science, Design, and Engineering of Contagious Ideas by Dan Zarrella

Dan Zarrella's an internet marketing Guru. He's researched internet marketing and social media, and has some ideas about what makes things go viral. Some of them are encapsulated in this book.

The book is very small and very short. Much of it was previously posted on Dan's web site. However a small book-shaped package is a convenient format. Dan says the bunnies on the front represent rapidly-reproducing ideas.

Dan uses memetics, cites Richard Dawkins and says:

Our world is made of memes. If you've ever seen the matrix movies, You'll remember their world was composed entirely of computer code. Everything people interacted with was built from computerized instructions. Similarly, our world is made of contagious ideas. Everything made by huamns - from the chair you're sitting on, to the book you're reading - exists only because someone had the idea to invent it and that idea caught on, spreading from person to person.
It's a memorable image: our world is indeed made of memes.

The book is full of social media marketing tips of the type Dan posts on his blog. It's full of graphs and charts telling you what and when to tweet for the best results.

There was one bit of the book which I really didn't like - where Dan defined a measure of the rate of increase of memetic infections per generation, claimed that trying for an explosive epidemic was unrealistic and then recommended using big seeds.

Dan doesn't seem to think small seeds are effective. It is true that you should spend some of your marketing budget on seeding your idea. However making your idea spreadable is really very important. Dan says that when you do get a viral idea, it's just a fluke, and you shouldn't build your marketing strategy on luck. But relying on big seeds is not really correct advice in general. Pop songs may not reach every single member of the population before dying away, but they do reach many millions and that's good enough for their composers.

Some do have to rely on big seeds, since they have content that requires it - but most should try and use highly-contagious memes in their marketing, for best effect.

Dan advocates a science of marketing. However, few marketers do very much science, since they often don't want to publish their raw data, and they often don't trust what other marketers say. It's probably more realistic to advise marketers to cherry pick the best bits from the scientific method - such as iterating the process of performing experiments, measuring their outcomes and making changes. Maybe it's best to regard marketing as a technology - rather than a science.

The book is pretty neat. It is short, readable and fun. Most readers will probably be hungry for more details, but at least this is a start.


Sunday, 2 December 2012

Tim Tyler: Pagel, Wired for Culture (review)


Hi. I'm Tim Tyler, and this is a review of this book:

Wired for Culture by Mark Pagel.

The book says it is about how and why humans have evolved so that they were able to rapidly build today's large-scale, complex societies. It discusses cooperation, deception, religion, altruism, kin selection, reciprocity, reputations and language. This is an interesting set of topics and the book offers a unique perspective on them.

The book is largely about origins - so a hefty chunk of it is about things that happened thousands or millions of years ago. It is written more as a popular science book than a science book. The book was definitely a good read, but alas, I often found myself irritated, frustrated or in disagreement with the author while reading it. I'll start with some of the negative points, and then get on to what I liked.

The book is saturated with human exceptionalism. We hear that only humans have proper culture and a proper language. Mark discusses cultures and languages in non-human animals, but is ultimately dismissive of them, since non-human cultures are insufficiently cumulative, and non-human languages are insufficiently symbolic. I tend to emphasize the opposite perspective. I think it is important to see the close links between human and non-human animal cultures and languages and use comparative ethology to illuminate these features of human social life.

Next: cultural evolution. Mark is aware of cultural evolution, and it plays an significant role in his narrative. At the very start of the book, he discusses the subject in terms of memes, gives Richard Dawkins credit for the idea and then lays out his understanding of the topic. Mark cites Dennet, discusses brain flukes, rabies, and the Cordyceps fungus. Wilson's idea of genes holding "culture on a leash" is discussed and Mark invokes the Terminator, and HAL (from 2001) to illustrate how Wilson might turn out to be completely wrong about that. He discusses memes that produce suicidal behaviour, compares such negative memes to viruses and then discusses the idea of a cognitive immune system with its own internal Darwinian mechanisms whose function is to reject bad memes. This part of the book is quite good.

However, cultural evolution is not really the topic of the book. Indeed, it receives only occasional attention in the rest of the book. Instead, the book is all about human evolution in terms of DNA genes. Mark emphasises that culture is generally good for our genes, and generally treats culture as a set of tools that genes have used to get what they want.

Many of those discussing the origins of human ultrasociality invoke gene-meme coevolution, with memes driving and genes being dragged around in their wake. Mark does invoke lactose tolerance as an example of this effect, but this kind of scenario doesn't really feature prominently in his narrative. Indeed, in most cases, where I would invoke benefits to memes, Mark instead talks about benefit to genes - sometimes mentioning the idea that culture is generally good for us, or we wouldn't have it. There's quite a large literature on how genes and memes coevolved to produce the human social mind, and Mark hardly cites any of it. There are a few jabs at "strong reciprocity" and "group selection", but that's about it. Mark obviously has a strong scientific background, and he does have some understanding of cultural evolution - but I was left with the impression that he hadn't finished thinking through its implications, and hadn't read much of the literature on the topic.

The story of human evolution is really the story of the rise of memes. That story is far more interesting than the tale of the bigger brain and the modified vocal chords. By concentrating so much on DNA genes, Mark misses out most of this story. Cultural evolution is actually very important to understanding how human DNA evolved, due to meme-gene coevolution. For example, once people start to consider cultural evolution properly their story about why humans have big brains typically starts to look very different from the "Machiavellian intelligence hypothesis" that Mark advocates in this book. For me, the emphasis on DNA genes was the single biggest problem with the book.

Mark discusses the idea of "cultural survival vehicles" - using terminology that Richard Dawkins once attempted to bury. The idea of what counts as a cultural "vehicle" broadly corresponds to the idea of what a cultural 'organism' is. Organisms have poorly delimited boundaries in the organic realm - what with ant colonies, slime molds, and symbiotic unions - such as the Portugese man'o'war. However, the situation in cultural evolution is even muddier, with transient coalitions between memes arising on many different scales. CDs, MP3s, PDFs are examples of things that many would treat as cultural organisms. However, Mark doesn't really discuss the whole issue of what counts as a cultural "vehicle". Instead he defines "cultural survival vehicles" to be the cultures associated with tribes of humans, and then uses this definition consistently throughout the book. My assessment is that Mark's approach to the topic of what counts as a "vehicle" in cultural evolution was narrow, dogmagtic and poorly thought through.

Next Mark's treatment of group selection. Mark devotes a few pages to criticism of group selection in the book. However, he doesn't seem to me to have tracked the recent literature on group selection very well. For many years group selection enthusiasts believed that they had a new theory that acted as a superset of kin selection. However after many attempts to say exactly what it was that group selection predicted which kin selection did not, the group selection enthusiasts mostly seem to have given up on this, and now largely recognise that modern group selection and kin selection theories make the same predictions - and so represent different ways of looking at the same process. Mark embraces kin selection, but is sceptical of group selection - a position that doesn't make sense if these are ideas that make the same set of predictions - as most modern group selection advocates now agree is true. I expect modern group selection advocates won't be too impressed by Mark's criticisms. They will just say that he failed to understand their position. As far as I can see, they will be correct.

On page 81, Mark attempts to explain patriotism and nationalism in terms of a "special" and "limited" form of nepotism by which a single gene for altruism recognises itself in other individuals and then helps them. Mark's explanation of this phenomenon appears to be unorthodox to me. He offers no references to supporting scientific literature. I don't think the explanation Mark offers for the existence of patriotism and nationalism is correct.

Mark also offers an explanation of how individual cells in slime mold populations can help each other despite being unrelated to one another. However Mark's explanation seems largely unnecessary to me - since cells in slime mold populations which exhibit multicellular stages are almost always close relatives - a fact which Mark fails to mention.

I thought that there quite a few of these dud explanations in the book. The lack of references to supporting material and uniform authoratative style made it hard to distinguish to good explanations from the bad ones. I wound up not entirely trusting a lot of what Mark was saying, and making mental notes to check up on his facts in cases where he was presenting material I was not familiar with. That is not a great relationship for a science writer to have with the reader.

Much of the book is about the reasons why humans cooperate. However, this material is distributed over many chapters and covered in a rather rambling way. The book really needs a summary of this material, saying which mechanisms are important.

Mark's most frequent answer to the puzzle of cooperation is that it pays. This is true of reciprocity and reputations, and it's true with the forms of byproduct mutualism which are frequently mentioned in the book. The coverage of cooperation in the book deals reasonably well with the theories of kin selection, reciprocity and reputations - though the cultural versions of these theories receive rather cursory treatment. However the book had little coverage of the important topic of cooperation due to manipulation. I remember one section on the subject - about how people used religion to manipulate each other. Memes manipulate humans into interacting with one another because contact between humans facilitates their own spread - but Mark doesn't discuss such possibilities.

Towards the end the book has some rather rambling digressions. Genomic impriniting may be fascinating, but it doesn't seem to have a lot to do with the main topic of the book. Mark also discusses his research on human hairlessness as an anti-parasite adaptation. An interesting topic no doubt - but again, not a whole lot to do with the main theme.

While it refers to the relevant science there are very few citations of actual papers. Instead the work of scientists is spun into a narrative. It's a narrative that is pretty well written. The lack of references in the body of the text probably helps a little with the flow, but at some expense to the science. There are references for each chapter at the end, but inconveniently, there are no footnotes or endnotes to explain how they relate to the chapter.

Enough about the dubious aspects of the book. There's also a fair amount of good material in it:

The book contains a pioneering section on cultural kin selection. Mark offers an explanation for altruism in humans that invokes the green beard effect - and makes it clear that he thinks that the explanation applies to both genes and memes. He uses terms such as "cultural relatedness" and "cultural nepotism". He also correctly explains that the green beard effect is just kin selection applied to individual genes or memes. This material is good. However, there are no references and no attempt to place this theory in its historical context. Nor is it a very comprehensive treatment of the subject. I would have preferred this section of the book to be longer. I'm rather sceptical that the green beard effect is the best way of describing cultural kin selection. We don't describe organic kin selection in terms of the net effect of a bunch of green beard genes - and I don't see any terribly good reason for describing cultural kin selection in those terms.

Mark offers a fairly withering critique of use of the ultimatum game by some researchers to illustrate how humans cooperate in anonymous one shot interactions where reputations and reciprocity can't possibly be involved - and so therefore the observed cooperation must be down to group selection. Mark offers what seems to me to be the obvious refutation - that subjects tend to play it safe in case they are not really anonymous in the laboratory environment - an explanation which had been previously given by Andrew Delton in 2011. As Mark says, the ultimatum game represents poor quality evidence for kin or group selection.

In summary, this book is rather patchy, with some good bits and some bad bits. The good bits are good enough to make the book worth reading, though. It's certainly nice to have a professor of evolutionary biology using memetics and cultural kin selection.