processPhotography - Networks > Clusters > Links

Network thinking is poised to invade all domains of human activity and most fields of human inquiry. It is more than another helpful perspective or tool. Networks are by their very nature the fabric of most complex systems, and nodes and links deeply infuse all strategies aimed at approaching our interlocked universe. Albert-Laszlo Barabasi

../Images/green.png

A more recent field of inquiry that is advancing research about Complexity focuses on the structure of complex systems. The emerging science of networks examines and quantifies the behaviour of real world networks. "Networks are the prerequisite for describing any complex system, indicating that complexity must inevitably stand on the shoulders of network theory."_{Barabasi 2003 p238}

There have been some key advances by network science that allow us to better understand the networks that are at the core of our connected age. The way in which social networks form; the truth behind six degrees of separation; why technologies, products and people become successful; and how we understand history in light of these advances, are of particular interest to this project.

Network science explores both the dynamics of the network - the evolving structure; and the dynamics on the network - individuals doing something _{Watts 2003}. Networks represent populations of individual components or nodes, which can be anything from people to pages on the internet. Nodes actively link, forming connections, which in turn form clusters. The way in which connections and clusters form, and the type of networks and systems that result from these clusters are what interest researchers.

The study of the connections and clusters of real-world networks covers a vast array of test cases including the novel and the famous. It still comes as some surprise though, that in the world of network science Kevin Bacon has a starring role - although it's not because of his dancing in Footloose. Rather, it is because of a game developed by three college students in 1994, who set out to prove Kevin Bacon was at the centre of the Hollywood universe. The object of the game is to determine an actor's Bacon Number by figuring out the shortest path to Kevin Bacon. Kevin Bacon has a Bacon Number of 0, anyone who has acted in a movie with Kevin has a Bacon Number of 1, anyone who has acted with someone who has acted alongside Mr Bacon has a Bacon Number of 2 and so on. The game was developed further by Glen Wasson and Bret Tjaden in 1997 who linked the game to the largest online movie data base IMDB, and is now known as The Oracle of Kevin Bacon. By typing in the name of any actor, the shortest path is found between the actor and Kevin, listing the chain of actors and movies by which they are connected. The real challenge, however, is to find an actor with a high Bacon Number - since most actors in Hollywood can be connected to Kevin with two to three links.

Transcluded from http://oracleofbacon.org/

The Oracle of Kevin Bacon inspired British photographer Andy Gotts to try to reach Bacon through photographic links instead of film links. He wrote to 300 actors asking to take their photograph and received one response from Joss Ackland. Ackland suggested that Gotts photograph Greta Scacchi, with whom he had appeared in the film White Mischief. Gotts continued from there, seeking referrals from each actor he photographed. Eventually, Christian Slater referred him to Bacon. After eight years, Gotts's photograph of Bacon completed the project. Gotts has published a book called Degrees.

This image is a link to Andy Gotts Website

So is Kevin Bacon really that special?

One might conclude, as did the fraternity brothers, that there is something special about Mr Bacon, that he is somehow the fulcrum on which the universe of actors tips. But thinking about it for moment longer, an entirely different interpretation seems more plausible. If it is true that Bacon can be connected to almost anyone in only a few steps, might not it be true that anyone can be connected to anyone else in roughly the same number of steps?_{Watts 2003}

By studying the data from The Oracle of Kevin Bacon, Watts and Strogatz found that Hollywood is a small world network. Small world networks are highly connected by nature and as such it is relatively easy to connect two people through a short chain of acquaintances. It is also commonly known as six degrees of separation. In the case of Hollywood, "every actor can be connected to every other actor in an average of less than four steps"_{Watts 2003 p95}. More importantly, Watts and Strogatz as well as research by Barabasi, found that small worlds are a generic property of networks in general. "It is rooted in their structure", it just doesn't take that many links to connect two people_{Barabasi 2003 p40}. Unfortunately for Kevin Bacon Fans everywhere it dispels the notion that Kev is special :-( !

The Ties That Bind

Watts and Strogatz's research into small worlds, uncovered a number of significant properties of networks. They found the most important links in networks are 'weak ties' which "can be thought of as a link between individual- and group-level analysis in that they are created by individuals, but their presence affects the status and performance not just of the individuals who "own" them but of the entire group with they belong"._{Watts 2003 p49}

The importance of weak ties between individuals who don't know each other very well or share common interests, is that they are the source of effective coordination and information. With densely connected strong ties, the information shared is the same, "the weak ties, or acquaintances, are our bridge to the outside world, since by frequenting different places they obtain their information from different sources than our immediate friends"._{Barabasi 2003 p43}

They also found that the multiple linking or clustering that occurs in social networks is in fact a universal feature of networks in general. Barabasi points out that clustering is something we intuitively understand, we have an "inborn desire to form cliques and clusters that offer familiarity, safety and intimacy"_{Barabasi 2003 p49}. Barabasi's own research shows that hubs, which are highly connected nodes, are a fundamental property of most networks and form hierarchies in networks. "Hubs appear in most large complex networks. They dominate the structure of all the networks in which they are present, making them look like small worlds"._{Barabasi 2003 p64}

Rich Get Richer

Growth is an essential feature of most networks. The World Wide Web has over a billion documents and it is growing, node by node, page by page. With such rapid expansion, the way we choose to navigate the web is typically through highly connected hubs, such as Google or Yahoo!. As Barabasi points out:

When choosing between two pages, one with twice as many links as the other, about twice as many people link to the more connected page. While our individual choices are highly unpredictable, as a group we follow strict patterns._{Barabasi 2003 p85}

Barabasi shows that growth and preferential attachment are two fundamental mechanisms governing network evolution. Growth plays an important role in expanding the network and determining how nodes link:

If a node is the last to arrive, no other node has the opportunity to link to it; if a node is the first in the network, all subsequent nodes have a chance to link to it. Thus growth offers a clear advantage to the senior nodes, making them the richest links. _{Barabasi 2003 p87}

Furthermore, he shows that preferential attachment equates to the idea that popularity is attractive. New nodes prefer to link to popular, well connected nodes. Early nodes therefore will be selected more and will grow faster than newer less connected nodes.

As more and more nodes arrive and keep picking the more connected nodes to link to, the first nodes will inevitably break away from the pack, acquiring a very large number of links. They turn into hubs. Thus preferential attachment induces rich-get richer phenomenon. _{Barabasi 2003 p88}

In a competitive environment another factor is added to the dynamics of the network. Fitness is the measure of a nodes' ability to stay in front of the competition. Nodes with higher fitness are linked to more frequently. Therefore a latecomer with great technology can acquire links faster. "Beauty over age"._{Barabasi 2003 p97}

Cascading Success

How innovations take off, how people and products succeed, has also been the focus of network science. Watts has shown that the key is the way in which the new information is taken up by the network. "The structure of the network can have as great an influence on the success or failure of an innovation as the inherent appeal of the innovation itself"_{Watts 2003 p244}. There are instances, i.e. cultural fads or riots, where "individuals in a population essentially stop behaving like individuals and start to act more like a coherent mass"_{Watts 2003 p205}. The drive behind this emergent behaviour is known as an information cascade.

Information cascades are rare events and depend on many variables or externalities. However, what Watts finds is common to all information cascades, is that once one starts it becomes self-perpetuating - "it picks up new adherents largely on the strength of having previous ones"_{Watts 2003 p250}. It's Barabasi's growth and preferential attachment gone mad!

The difference between a highly successful innovation and an abject failure can be generated through the dynamics of the interactions between players who might have had nothing to do with its introduction._{Watts 2003 p250}

Furthermore, Watts shows that the ultimate success of a product, person or new idea, has less to do with the innovation, and more to do with the connectivity of the network. "The trick is to focus not on the stimulus itself but on the structure of the network that the stimulus hits"_{Watts 2003 p249}. It follows then, that in the case of an information cascade, the stimulus has a disproportionately large effect on a network. In other, more familiar words, the whole is different from the sum of its parts.

A Nonlinear View of History

Information cascades are unpredictable; they demonstrate "that inputs and outcomes cannot be associated in a proportional or even unique way". Indeterminacy, chance, and non-linearity are acknowledged as important aspects in Complexity. This understanding is, however, incongruous with conventional history which tends to promote order, and neglect the arbitrariness in the world. History also ignores the possibilities - the things that might have happened. Because of this, we tend to assume that the actual outcomes were preferred to all the alternatives.

What we don't generally consider is that the very same thing, with the very same characteristics, just as easily could have been a dismal failure. Nor do we typically lament the multitude of unsuccessful innovations that also could have been contenders had their circumstances been perhaps slightly different._{Watts 2003 p244-5}

Not only do we presume the actual outcomes are somehow preferred, we assume that the success of those outcomes are related to intrinsic "characteristics that happen to be exhibited"_{Watts 2003 p244} . This is despite the fact that those characteristics may not have been acknowledged as being particularly special beforehand. Network science shows us that success cannot be established before the fact.

Conventionally, when something or someone is successful, we assume that the extent of the success is proportional to some underlying measure of merit or significance. Successful artists are creative geniuses, successful leaders are visionaries; and successful products are just what consumers were looking for. Success, however, is a descriptor that can only be applied after the fact, and with hindsight it is easy to be wise._{Watts 2003 p244}

In light of complexity, qualities like greatness, are also re-evaluated. Assumed to be inherent in a person, product or idea - greatness is only issued after the fact - even though our perception is it was there all along. Before the fact, it is rarely clear what outcome any particular state of affairs will produce. Greatness, like success , is "a consensus arrived at by a large number of individuals" engaged in a multitude of dynamic and complex interaction. "Such contingent decision making comprises the essence of an information cascade, and in doing so renders the relationship between cause and ultimate effect deeply ambiguous"._{Watts 2003 p245-6}

Complexity illustrates that History is "an unreliable guide to an unpredictable future"_{Watts 2003 p302}. Complex systems and networks show us, that our dynamic interactions now, shape the possibilities and outcomes of our present and future. "The notion that outcomes can be properly understood in terms of the interactions of individuals, each of whom is reacting in real time to the decisions and actions of others, presents us with a quite different view of cause and effect than the one to which we are generally accustomed". Complexity redefines our view of history from the familiar linear-progressive narrative, to a view regulated by dynamic exchanges, indeterminacy, and non-linear relationships between cause and effect. Complexity offers photography a non-linear future.