Artwork from Takashi Murakami at the Hong Kong art fair 2010

Sailor Moon (1992 – 1997) transcended Japanese anime and reached a pure state of corrosiveness. Sailor Moon is an artificial flavouring substance: depthless, highly satisfying and addictive. It is more than any artwork of Takashi Murakami the best illustration of his superflat art movement, depicting “the shallow emptiness of Japanese consumer culture”. The original manga is a little different from the anime and somehow more spiritual. The anime version expunged its scenario of any particularity, leading to the ultimate stereotype of the Japanese girl, flanked with kitsch accessories ready for merchandising, cheap love stories and consumerist lifestyles. The characters were designed as for any animes to appeal both to girls and perverts thanks to a quota of ‘subliminal’ underwear scenes. Their transformations into self-centred wonder women are the climax of every episodes (otherwise rather mediocre in their drawings). The same scenes of transformations are shown again and again, becoming objects of cult, obsessing and hypnotic. They saturate the narrative with their superflat symbolic.

There is a before and an after Sailor Moon, the series imitated many other predecessors, and has been often imitated, but it was the first to reach that level of depthless consumerism worship.

Schelling’s segregation model for an open city: emergence of physical frontiers from a socio-spatial dynamics

By Laetitia Gauvin, Jean-Pierre Nadal and Jean Vannimenus

Segregation happens even if personal preferences to live closer to a community are small. In the 70s Schelling introduced a multi-agent model to describe the segregation dynamics that may occur with individuals having only weak preferences for ‘similar’ neighbours. Here we present a model where people move to a new town, which can be more or less attractive, and how they spread in the city. With a varying vacancy density, the leads to a variety of “phases” whose main characteristics are the interfaces between clusters of agents of different types (segregations). These interfaces may under certain condition partially or completely isolate groups from each other, leading to different phases of segregation. Here you can see what theses phases look like. Depending on the value of some parameters, the neighbourhoods are more or less clustered and dense, leading in some very specific circumstances to physical buffers between them.

Model of urban segregation

http://arxiv.org/abs/1002.3758

It is thus possible to model mathematically how different cultures and social classes cohabitate in a same environment. Will a multicultural environment be one day conceived at the urbanistic level thanks to complex systems theory?

This is what Alexy Khrabrov and George Cybenko attempted to answer at the ECCS 2010 conference in Lisbon. Here is an introduction that I have slightly edited to make it more accessible.

“Mind Economy: Modeling Influence in Communication Networks with Social Capital.”

by Alexy Khrabrov and George Cybenko

Social scientists, businesses, and governments are interested in summarizing the ongoing social network activity to identify the most influential players capable of creating and maintaining high-impact group behaviours. We would like to have metrics of influence in dynamic systems, and generative models which can explain how this influence is accumulated and maintained. Having identified the “stars” or high-influence individuals, we look at the ways they achieve and maintain their influence, comparing their tweeting behaviour to social capital exchange in proportion to the fans ‘contributions. We propose a family of generative models where social capital is exchanged and generated during interactions, reflecting the players‘utilities – such as self-centred or maximizing group benefit. Using our social capital model, we let a system evolve to accumulate most of the capital in those nodes which can be considered influential. We compare those capital-rich nodes with other metrics of influence and show that our model confirms and explains influence of many important types of players, and reveals the behaviours leading to sustained influence. We model recently uncovered Twitter phenomena such as Justin Bieber‘s ecosystem and other high-intensity processes, showing how efficient star behaviour and group preferences lead to various mind economies in social networks.

http://cs.wellesley.edu/~cs315/Papers/khrabrov-twitter-dynamics.pdf

So, even the seemingly chaotic activities of a teenager can be modelled using economic models in which influence and social capital is the currency. Trends on social networks such as twitter and facebook still seem rather unpredictable. But the interests at play in a mind economy are too big to let it go that way. Complex systems theory will undoubtedly be used in an attempt to better target influencers, and shift our attention to a particular subject. The same theory might on the other hand help us making sure that social networks maintain a certain level of social fairness in the mind economy, and diversity in mindshare.

Wikipedia and Linguistic Networks

(“Generating Linguistic Networks Based on Large Corpora of Linguistic Data”)

By Alexander Mehler

With the rise of the web, linguistic networks such as Wikipedia reach a size, structure and complexity that have been widely unknown so far. These networks induce a further level of information structuring above the level of textual aggregates and their constituents. For example, Wikipedia allows the classification of its pages into categories that are defined dynamically by the community. Alexander Mehler analysed these categories and showed some evidences of interesting properties. For example, that pages belonging to a same category can have characteristic text structures recognizable without the need to understand the meaning of the texts. In the same way, the structure that forms Wikipedia in different languages seems characteristic as well. Just by looking at the relations between the categories, one might have a clue of their language, without having to look at the texts! These methods are not accurate enough to be predictive, but they can increase the accuracy of semantic analysis. Findings like these uncover structures of our knowledge that were until now unknown or hypothetical.

Social influence model of language competition

By Wojciech Borkowski and Andrzej Nowak

Many languages that existed not long ago have either died, or are listed as endangered. Researchers concentrate on developing models that can reproduce the empirical distribution of language sizes. Social impact may be responsible for many phenomena known from the study of languages, such as the puzzling correlations between grammar features across different languages.

Poster: language competition model

So, the disappearance of cultures and languages might be no more than the consequence of a game of influence?

A Mathematical Approach to the Study of the United States Code.

By Daniel Martin Katz and Michael Bommarito

The United States Code is a document containing over 22 million words that represents a large and important source of Federal statutory law. Scholars and policy advocates often discuss the direction and magnitude of changes in various aspects of the Code. Does it get more complex with time? Daniel Martin Katz and Michael Bommarito formalized mathematically the notions behind this question, and demonstrated by the analysis of the system that the Code has grown from 2008 to 2010 in its amount of structure, interdependence, and language.

http://cdsweb.cern.ch/record/1254777/

Could one day legal texts be analysed scientifically in terms of logic and efficiency?

Simulating opinion dynamics in heterogeneous communication systems

By Walter Quattrociocchi, Rosaria Conte, and Elena Lodi

In this video, Rosaria Conte describes opinion dynamics by means of multi-agent based simulations. Agents (i.e. people) are exposed to different sources of information varying both the contents and the perceived reliability of the messages spread. Agents’ internal opinion is updated either by accessing one of the information sources, namely media and experts, or by exchanging information with one another. They are also endowed with cognitive mechanisms to accept, reject or partially consider the acquired information. The study evaluates the impact that reliable sources and peer-to-peer communication can have on the quality of information.

Simulating Opinion Dynamics in Heterogeneous Communication Systems from Assystcomplexity on Vimeo.

At the third minute of the video, Rosaria Conte starts attacking the Italian media manipulated by Berlusconi. Behind the rather tedious title, this presentation is a fantastic example of political engagement through science, which doesn’t undermine at all the scientific value of the research.

Opinion dynamics

By Anthony Woolcock

Many societies exhibit cultural fragmentation. This is despite individuals trying to reach agreement with those they meet. In the model of Axelrod, individuals that are more similar are more likely to interact (homophily). The mechanism where people become more similar after interaction is termed social influence. Axelrod’s model is interesting because for different parameter choices the opinions of all the individuals will either become all the same (consensus), or frozen fragmented state. A phase transition is observed between these two types of frozen state.

Complex systems theory: opinion dynamics

Bounded confidence model: addressed information maintain diversity of opinion

By Krzysztof Malarz and Krzysztof Kulakowski

Models have already been developed to optimize the frequency of let’s say political advertising on tv in order to get the maximum effect on a public’s opinion (see the Zaller mass opinion model). But they didn’t take into account interpersonal communication, which becomes more and more decisive with the advance of online social media. Krzysztof Malarz and Krzysztof Kulakowski developed a model taking this into account. Their model suggests that individually addressed messages maintain diversity of opinion. See the full paper here.

Model for diversity of opinion

I don’t know many fields that are not confronted with complex systems. Chemistry, physics, maths and all the scientific disciplines I studied at university are dealing with structures that require more than simply understanding their parts. I had the occasion to deal with other types of systems when I was working in business innovation: organisations are made of networks of people, networks of processes and are intricately connected with their economical ecosystem. I have seen many complex systems in culture too: semiotics, linguistics and urbanism just to give few examples. Finally, in the researches I’m currently making on consciousness, systems play also a major part, such as in neuroscience. So, complex systems are everywhere. We are only beginning to understand their mechanisms, and how interdisciplinary their study is. I will publish a series of posts on the subject in the coming weeks, which I hope will interest you as much as I was by the conference.

The presentation of Professor Uri Hasson was fascinating (see this article from the New Scientist). By using fMRI and observing brain’s activity, it is now possible to tell if two spectators go through similar mental states when watching a movie: do they pay attention to the same things (e.g. a face or a decor); do they experience the same picks?

Uri Hasson was careful to not suggest that movies controlling more effectively brain’s activity are necessary better (ambivalence can also be a virtue). He also emphasized how his researches could be used for other purposes than studying films. Films could be used to simulate the effect of real-life situations on the brain for example.

However, it is very easy to predict what Hollywood studios will make of these researches, whatever their initial purpose was. They will use them to optimize trailers and films so that they can generate the ‘optimum’ effect on the brains of the audience. They already do so, but through approximate methods, such as asking a test audience to explain what they experienced when watching a movie (technique used for Harry Potter films). Hollywood studios have always been obsessed in controlling their audience, for economic reasons, or American propaganda. It is very frightening to think of what they will do with this new technique.

The film director Amos Gitai didn’t let it pass. He overreacted to the research, accused Uri Hasson of having a hidden agenda, and clearly wished that the research never existed. He felt that his artistic practice of filmmaking was threatened. The debate became very passionate and confrontational between the two orators and the audience, in a typical German fashion.

I think for my part that we cannot ignore the evolution of science. It is counterproductive to simply wish that it didn’t exist, especially when the research could also have positive applications. Hollywood studios will use any trick at their disposal to ‘control’ their audience. The audience should be aware of these techniques; this is why this talk is so valuable. The audience has also the capacity to choose other options, and to support instead independent filmmakers and alternative movie theatres, which are becoming increasingly popular.  Even when they watch a Hollywood blockbuster, and I do sometimes, they can let their mental states being controlled for a while, and still remain critical. The problem lies more in education and the audience’s capacity to remain critical, which is sadly appalling in many socio-cultural groups.

If you are also interested in subjective cartography, I recommend these strange maps.