Creating harmony in complex networks

Scientists working with musicians have revealed new strategies to deal with frustration and achieve synchrony in complex networks.

In fact, they say, they’ve found that humans working together can achieve something that other networks can’t – and that includes birds, fish, planets, particles and even pendulums.

“I found it surprising that the ability of humans to ignore frustrated signals changes completely the dynamics of human networks,” says Moti Fridman from Israel’s Bar-Ilan University, senior author of a paper in the journal Nature Communications.

And this, he adds, enables humans to find better solutions than current models have predicted.

Although we know humans can ignore distractions, like having a conversation in a crowded place, network models of human synchronisation hadn’t considered this until now – and this has many important ramifications, according to the researchers.

“Synchronisation is crucial for the life of all living species on our planet, from the cellular level to the crowd synchrony of large groups,” they write.

“In particular, the synchronisation of human networks is essential for our civilisation and can impact the physical and mental wellbeing of individuals in groups.”

The applications, they say, extend to managing networks as diverse as traffic, economics, epidemics and decision-making processes in different groups.

Essentially, Fridman says, it relates to any network where each node has decision making power, including artificial intelligence and autonomous cars. “Our model can predict with high accuracy the dynamic of such systems, beyond what was possible before.”

His research initially explored how lasers and even singing wine glasses synchronise and he wanted to try it in humans, but how to do this eluded him – until he met musician and PhD candidate Elad Shniderman from Stony Brook University, US, a co-author on the paper.

At first, they set out to demonstrate how existing models work, and chose violinists because the public can see and hear the synchronisations.

The surprise came when they analysed the initial results, which revealed previously unidentified physics of synchronisation that they verified through several replications and configurations of the experiment.

Current models look at all-to-all coupling, or global coupling, in which all players would hear everyone, similar to a crowd synchronising their clapping, and have had little control over factors such as network connectivity, coupling strength and background noise.

In a series of experiments, Fridman and colleagues addressed these limitations, creating a complex coupling network in which each player is connected to a small number of players who are also connected to other players.

Sixteen isolated violinists played a recurrent musical phrase, which was recorded and played to them via noise-cancelling headphones in various network formations – this was the only input as they couldn’t see or hear each other.

A computer-controlled system created varying delays on the coupling between them, and they were instructed to synchronise their rhythm to what they could hear.

The players showed several strategies for finding a stable solution, such as ignoring conflicting signals and effectively deleting connections if another player was playing at a different tempo to the rest of the group.

It’s not an experiment in individual human psychology, Fridman notes – it’s about the group. “Our results predict the global behaviour of the network; we don’t know how each player will choose, but we can say a lot about the dynamic of the entire network.”