Butterflies in the Net
Luís Ángel Fernández Hermana - @luisangelfh
19 June, 2018
Fecha de publicación original: 25 abril, 2000
One grain does not fill a silo, but it helps its companion
Someone knocks on your door and it’s a total stranger who says, “Hello I’m Australian. My cousin such and such travelled with a girl who was married to Tom, who knows someone who studied with Dick, who is, I believe, a good friend of yours.” You have two options. Either you say, “Come in and make yourself comfortable, have a whiskey”. Or, otherwise, you bark bad-temperedly, “I’m not interested in encyclopaedias”, and slam the door. Reactions to this miniscule incident, which all of us have experienced at some time or another in our lives, actively or passively (I admit that I have been both perpetrator and “victim” many times – the inevitable consequence of thousands of miles of hitch-hiking) are fundamentally important. Firstly, if we say yes, come in and tell me all about your life, we have no idea just how widely we open up to the world thereby discovering just how small it is. If we say no, go away, we also have no idea as to just how surprisingly many doors we have just closed. This is all just simple common sense. However, it is also hard science as Duncan Watts has demonstrated.
US researcher, David Watts, decided to test a popular saying in his country: everyone is just six handshakes away from the President. Is this really true, he asked, or just a popular myth perpetuated to show that apparently inaccessible people are really close to us? Is the world really so small or do we just exaggerate the most trivial coincidences in our heads? To solve this problem, Watts used advanced mathematics, super-computation, and all the most modern weapons in the statistical arsenal. And he reached the conclusion that, yes, the world is, indeed, a small place when we construct it with certain types of networks. The results of his work have become seminal in a variety of apparently unconnected fields such as social psychology, biology, epidemiology, the functioning of the brain or that well-worn concept of globalisation which in the light of his work takes on interesting overtones. Basically, everything comes down to how networks, of whatever kind, are constructed and behave. At the same time, in the case of human networks, this decides how close we get to each.
In his book titled “Small Worlds” (*), Watts explains that his point of departure was a simple question: What is the real dimension of the world from the point of view of possible contact amongst its inhabitants? His research, of course, rapidly led him to complexity. It is not easy to make models of networks in a laboratory, nor apply them to the real world to ascertain their exactitude.
Nevertheless, he came up with some surprising answers which, like a tin-opener, began to open up some knowledge areas hitherto the domain of intuition or statistical correlations of a dubious nature.
Watts discovered something that, to a certain extent, we already know: in very tightly closed, regular, structured networks, many steps have to be taken for one individual to make contact with another from another circle (“I’m not interested in encyclopaedias”). Nevertheless, when a certain amount of chaos was injected into the network, the number of steps became considerably reduced (“Come in and tell me all about yourself”). The most interesting revelation was that when a certain intangible critical mass of chaos was reached, distances, to put it into simple words, became practically neighbourly. To put it more graphically, all the individuals in complete, complex networks found themselves within arms’ reach. The network globalised.
Although human networks do not lend themselves so easily to examination as computer operated simulators, Watts selected some examples that confirmed the laboratory results. The density of contacts, their random nature, the simultaneousness of the access to the network by the users (anyone, from anywhere), their interconnection with other networks, exponentially increased the probability of people meeting and exchanges taking place amongst users. The network, as metaphor of the world, effectively became the size of a handkerchief, in fact a pocket-sized one. What was unusual here was that the encounters at all levels did not take place.
Watts’ work has cleared the way for a wide variety of disciplines which depended on networked behaviour. Epidemiologists, for example, are using his conclusions to predict, as far as is possible, the propagation of viruses, their scope and speed. Once they have a model, it can be tested in real life cases such as the AIDS and Ebola viruses. This way the saying that the beating of a butterfly’s wing in Japan can produce a tornado in Florida can become more than just a beautiful metaphor. An insignificant action in one place could become a phenomenon with enormous implications if it encounters dense and random networks with a high degree of potential for amplification. And this goes as much for networked communication –the Internet and others– as for the functioning of the brain where the neuronal connections, their proximity and density, play a fundamental role in moving and relating information.
The same could be said when applied to a networked-based economy, the internal organisation of a company, the workings of the market or the activities of inhabitants within the network, whatever these may be. Watts points out something that we have often noted from an empirical point of view when analysing the Internet: the key is understanding that the networks that make our world smaller operate without organised centres and by global interaction. Any individual in a networked community with the sufficient amount of randomness can access, at least in principle, all the individuals that populate it and interact with them. And the shorter and more local the connections, the richer, more extensive and wide ranging these are.
This is another way of expressing the relationship between the global and the local, and another way of demystifying attempts to heirarchify the Net based on different parameters, such as the economic potential of some of its members or their supposed capacity for imposing rules of conduct on the Net as a whole. In reality, we should view the Net from another perspective, namely, that the important thing is that each and every one of its users always retains the ability to interact within it, because it is that action that will modify it. And here we come to a factor that is often ignored when it comes to “thinking about” the Net: its continued growth does not add only quantities in demographic terms (which leads to the illusion of generating mass poles of attraction), but also qualities in terms of behaviour. For this reason, what is valid at one point in time varies constantly because every second thousands of butterflies are entering the Net and their wing beats could cause uncontrollable disturbances. In other words, the same unit of information changes due to factors external to it, such as the number of people interacting with it.
The other implications of Watts’ work already form part of Net culture. The model of heirarchical social organisation that has predominated up until now is bound to clash with that which is emerging as a result of virtual logic, that which belongs to a world of networks organised by interconnected computers within an open architecture, where none is more important than the other.
Heirarchical networks where the classical “I order, I command” applies, are closed, rigid, unfavourable to a multiplicity of encounters and guided by the prejudices of the large and the small, the important and the insignificant. Consequently, they lose out on the specific richness of globalisation and waste the added value of exchanges between those that have not even met face to face, but who have a whole lot to say to one another and lots of experience to contribute to the Net as a whole, which is what is logically prized these days.
(*) Duncan J. Watts, “Small worlds”, Princeton University Press; ISBN: 0691005419.
Translation: Bridget King