Networks can reveal complex relations between things, people, words and ideas. We can use networks to explore the relations, connections and associations between things in our research.
Visualisations can show us what artistic networks look like, from the outside, as a whole. They can convey information about the position of artists in relation to each other and the nature of their relationships – when did they arrive in the network, how long were they active, who did they work with, how often, over how long, and so on.
The Network interface provides tools to undertake and substantiate analyses of artistic netorks in live performance which were previously difficult to deliver. Researchers are particularly interested in using this interface to trace lines of artistic contact, influence and cross-fertilization and recognise patterns in career pathways and professional development.
Of course, visualisations can't tell us anything about the substance of artistic relationships. But for those with knowledge of the artists and productions that the network represents, network visualisation can provoke new ways of looking at historical evidence and new modes of story-telling and knowledge formation.
Some of the founding ideas about social networks were elaborated in the 1930s by Moreno. His research and practice focused on the connection between psychological well-being and the patterning of social relations between people. A key innovation introduced by Moreno in 1933 was the 'sociogram', a technique for depicting individuals as 'points' and their relations as connecting 'lines'. Moreno used the sociogram to envisage how small-scale networks of kinship, friendship, the workplace and so forth aggregate into larger networks such as communities, corporations, and nation-states.
Moreno's interest in networks of social relations has parallels in the performing arts. Consider the biographies in theatre programs, where the cast and crew each recount their relations with training institutions, theatre companies, leading artists and notable productions. A history of working where, with whom, and on what is how practitioners in the performing arts accumulate and articulate their experience, qualification and distinction. The interactions between artists, as they train, rehearse and work together, have implications for the kinds of artists they become and the kind of performances they make.
The impetus to visualise performing arts data emerges from the accumulation of data in AusStage. There is now, quite simply, more data in AusStage than it is possible to conveniently view on screen as text. Visualisations of various kinds – including charts, maps, timelines and network graphs – lend us the synoptic power of abstraction: they provide an opportunity to grasp the whole picture, to see relationships at a distance, where previously we had focused up-close on textual details.
With its relational design, the AusStage database records these networks of contact and collaboration and how they change over time. We are now able to analyse these networks by visualising data from AusStage and exploring 'who works with whom' in an online interface.
|Networks show the relations between things.
Here is something – it could be a person, an event, a book... or anything you like.
|Here are some more things.
The things that make up a network are referred to as nodes; nodes are also sometimes called vertices or points.
|Here are some relations between things.
The relations in a network graph are referred to as links; links are sometimes called edges or lines.
|How many relations does each thing have?
The number of links that a node has with other nodes is called 'degree of connection'.
|How far from one node to another?
The number of links between one node and another is called 'distance'.
|How strong is the connection between things?
The strength of the link is called the 'weight'. Weight can represent the strength, intensity, frequency or duration of relations.
|In what direction does the relation go?
Is the 'direction' of the relation one-way, like unrequited love? Or does it go both two-ways, like mutual admiration?
|Where in the network – at the centre or the edge?
This is called 'centrality' and there are several different ways to measure it.
|Do some things go together?
Things that are joined more to each other than to other things form 'clusters'.
This visualisation depicts a core sample from more than 1.6 million relationships between artists in the AusStage dataset. Analysing the whole network we learn that there are a few people who are highly connected, and a lot of people with a few connections. The average degree of connection is 42.73, the modal (or most common) degree of connection is 7. The network is very well interconnected. The average shortest path is 3.64 – which means that whereever an artist is in the network chances are that he or she can be linked to any other artist in less than 4 steps.
Examining the network for clusters of artists who frequently work with each other reveals something fundamental about performance: Theatre is made in the here and now – and therefore artists tend to work together in clusters, defined by time and place. What connects clusters across time and space are the writers – the playwrights, composers, lyricists, translators, choreographers and so on – who are often absent from rehearsal, but who contribute scripts and scores to the production from another time and place.
In visualising networks in AusStage, we focus on relationships between contributors and events. A contributor is an individual, usually a person, who contributes in some capacity to the conception, production or presentation of an event. An event is a distinct happening defined by title, date/s and venue; typically, a performance or series of performances at a venue. There are two ways we can look at the relationship between contributors and events.
|Contributor network – We can visualise a network of contributors linked by the events they have collaborated on. A contributor network is centred on one contributor. It shows all the contributors who have collaborated with the central contributor, and all of their collaborations with each other. Relations in a contributor network are mutual: if I've worked with you, you've worked with me.||Event network – Alternatively we can visualise a network of events linked by their contributors (below right). An event network is centered on one event. It shows the events that contributors were working on before and after the central event. Relations in event networks are directed in time, and show an artist's progression from one event to the next.|
In designing the network interface for AusStage, we were faced with a series of challenges. How to present researchers with a visual interface for exploring this dataset? Researchers access AusStage through a web browser. How much of the network could we deliver to them online? And how would they navigate through it? Through the course of exploratory research, a set of principles emerged which have guided the development of the network interface.
Our research on analysing and visualising the networks recorded in the AusStage database has developed in conjunction with the dissemination of scientific research on networks, complexity and emergence. Watts (1999, 2003), Ball (2004), Barabasi (2003), and Newman, Barabasi and Watts (2006) provide useful introductions to the field, and Borner et al (2007) and Golden et al (2009) provide up-to-date surveys. Statistical concepts for social network analysis are introduced in Scott (2000).
Ball, Philip. 2004. Critical Mass: How One Thing Leads to Another. Heinemann/Farrar, Straus & Giroux.
Barabasi, Albert-Laszlo. 2003. Linked: How Everything Is Connected to Everything Else and What It Means, Plume.
Bollen, J., Harvey, N., Holledge, J., McGillivray, G. (2009) 'AusStage: eResearch in the Performing Arts', Australasian Drama Studies, 54.
Borner, Katy, Sanyal, Soma and Vespignani, Alessandro. 2007. Network Science. In Blaise Cronin (Ed.), ARIST, Information Today, Inc./American Society for Information Science and Technology, Medford, NJ, Volume 41, Chapter 12, pp. 537-607. http://ivl.slis.indiana.edu/km/pub/2007-borner-arist.pdf
Goldenberg, A., Zheng, A.X., Fienberg, S.E., & Airoldi E.M. 2009. A survey of statistical network models. Foundations and Trends in Machine Learning, 2, 129-233. http://arxiv.org/abs/0912.5410/
Heer, Jeffrey and danah boyd. 2005. Vizster: Visualizing Online Social Networks, Proc. IEEE Symposium on Information Visualization (InfoVis), pp. 32-39, Oct 2005. http://vis.berkeley.edu/papers/vizster/
Moreno, J.L. 1953. Who Shall Survive? Foundations of Sociometry, Group Psychotherapy and Sociodrama. Beacon, N. Y.: Beacon House.
Moretti, Franco. 2005. Graphs, Maps, Trees: Abstract Models for a Literary History, London & New York: Verso.
Newman, Mark, Albert-Laszlo Barabasi and Duncan J. Watts. 2006. The Structure and Dynamics of Networks, Princeton, NJ: Princeton University Press.
Scott, John. 2000. Social Network Analysis: A Handbook, London: Sage Publications.
Walker, B.H. & D. Salt. 2006. Resilience Thinking: Sustaining Ecosystems and People in a Changing World, Washington, D.C.: Island Press.
Watts, Duncan J. 2004. Six Degrees. The Science of a Connected Age, Norton.
Watts, Duncan J. 1999. Small Worlds: The Dynamics of Networks between Order and Randomness (Princeton Studies in Complexity), Princeton, NJ: Princeton University Press.