Heroes Not Zombies

Smart World starts by acknowledging the work of two others – Albert-Laszlo Barabasi and Andy Clark. I’ve just read Linked by Barabasi. (ISBN 0-452-28439-2) It’s a fascinating book about the rather young science of networks. I agree with the author, that understanding how networks are created and function is going to be absolutely key to our future direction in science.

A network, quite simply, is made up of nodes and links. One example is social networks. Think of a piece of paper with the names of several individuals on it and lines drawn between the names of people who know each other. It’s remarkable how quickly information can spread around such a network. Maybe you came across the movie “Six Degrees of Separation” – a story based on the premise that there are only an average of six links between any two human beings on the planet. Turns out that idea, which apparently came from a Hungarian short story, is pretty accurate. But there’s a twist…….sometimes the number of links is way less than six (even between people who don’t know each other). Other kinds of networks you are familiar with are the maps of flight routes you see published in airline magazines, the power grid, and, yes, our dear World Wide Web. In fact, everywhere you look, you’ll see networks. Everything is connected. Nothing exists in isolation.

To try and understand how networks develop and how they function, Barabasi takes you on a journey through the world of mathematicians, physicists, social scientists and engineers. It’s quite fascinating. In the process he describes a very clear evolution of this new science. Intially, complex networks were thought to be completely random. But randomly created networks produced by computer modeling turn out not look like real world networks. Real world networks don’t have random distribution of nodes. Some nodes are way more connected than others. Barabasi calls these hubs. Once you introduce the concept of hubs, the mathematical modeling of networks reveal what are known as “power laws” (this is a bit beyond me I’m afraid – maybe Phil can help explain these?) but, as I understand it, if you take a single quality or characteristic in nature, say, height of individual human beings, you’ll get a bell curve. Bell curves look symmetrical and they have steep sides ie there aren’t many “outliers”. Complex, natural networks however have node distributions which can’t be described by bell curves. Instead you get a small number of highly connected nodes (hubs) and a huge number of less connected ones. This characteristic produces incredibly resilient and fast networks.

Real life networks are highly resistant to damage and they adapt to change. You can take out lots of nodes and not make much difference to the functioning. To really damage them you have to go for the hubs. Take them out and you bring the system down catastrophically. So, the structure of networks provides both their greatest strength and their greatest weakness.

Barabasi gives masses of great examples, from epidemiological spread of viruses like HIV, to the functioning of international economic markets, to the spread of ideas throughout civilisations. But one of his most interesting analyses is his critique genetics.

How often do you read about “breakthroughs” in mapping the genetic “origins” of various diseases – all with the promise of predictive genetic tests and of treatments based on what is known as pharmacogenomics – finding which genetic precursors determine the responses to which particular drugs. He dismantles this reductionist view very effectively and promotes a network model instead – making what I find to be a convincing argument that the genetic bases of diseases won’t be found in mapping the genome but in mapping the networks of genes.

This shift in perspective is crucial. It drives us away from a reductionist consideration of elements and parts towards a holistic consideration of system function by understanding nodes and their connections. He even terms this “postgenomic biology”. I like it! However, it’s at this point that he suddenly disappoints. His chapter 13 is very odd. It’s entitled “Map of Life” and in it he takes this idea of postgenomic biology and applies it in a bizarrely reductionist way, predicting that the future of medicine will be in tests and highly individualised drugs based on eliciting these genetic maps. He thinks you won’t need consultations with doctors any more, just simple blood tests which will be computer analysed and targetted, tailored drugs will then be kind of published on demand and delivered to your door and, voila! you have your own special cure!  I’m sorry, but I don’t buy this. I mean, I believe that if we could produce a new generation of highly specific drugs rather than the blunderbust ones we use now that would be great, but what happened to this idea of the science of networks, and how they would change our understanding of everything? Suddenly Barabasi leaps into a reductionist model of disease and healing which is predicated on the idea that each individual is indeed an island. Hasn’t he just spent the rest of the book showing us the importance of mapping connections? Isn’t every individual in fact massively connected not only to other individuals but to all kinds of environments. Isn’t it impossible to understand an individual as context-free?

However, don’t let chapter 13 put you off. He really is onto something extremely important here. Once you start to think this way you see networks everywhere and you begin to understand the inescapable importance of connections, and, interestingly, of hubs. We’re at the beginning of this science and I think it’s pretty exciting.

Those of you who have read other posts on this blog will be familiar with my references to Deleuze. His philosophy of networks – he preferred the model of the rhizome – predates this scientific development and has probably been one of the important nodes from which this area of study has grown. You’ll also be familiar with the concept of the Complex Adaptive System which I believe is the best model we have so far for understanding human health and illness.