Analysis: Predicting change: Scientists enjoy fertile times

Reuters News
Posted: Mar 21, 2011 4:18 PM

By Kate Kelland, Health and Science Correspondent

LONDON (Reuters) - Times of change can be drastic -- revolutions topple dictators, extreme weather kills tens of thousands and market crashes plunge people into poverty -- but for scientists studying complex systems they are fertile ground.

Few would want to predict the outcome of uprisings in the Middle East and North Africa, or of financial crises or natural disasters, but scientists say there are common traits in seemingly diverse systems that point to "tipping points."

Identifying and analyzing these, they say, can help people see the risks inherent in an interdependent world more clearly and take action to mitigate them.

"Complex systems can include societies, financial markets ecosystems like lakes or coral reefs, and even the brain," said Marten Scheffer of the University of Wageningen in The Netherlands.

"The fascinating thing is that while all of these systems have very different details and are very difficult to understand in some ways, they also all have one property in common -- that they can have tipping points."

The idea is to predict critical transitions in these complex systems when there is a sudden major shift to what scientists refer to as "an alternative steady state."

According to Yaneer Bar-Yam, director of the New England Complex Systems Institute in the United States, this kind of science is a vital tool for protecting societies from such dangers as pandemics, natural disasters, terrorism, climate change, resource exhaustion and economic crises.

His previous work has included looking at financial market crashes and building predictive mathematical models of ethnic conflicts in India, central Asia, the former Yugoslavia and Kenya, but he says it's not just about forecasts.

"Prediction tells you that something is going to happen, and that can be helpful so you can run away from it if you need to," he said in a telephone interview. "But much better is understanding why it happens, that can enable us to take action that may prevent things."


The key to understanding critical points such as leadership shifts or financial market crashes is to focus on collective behavior and not to analyze what an individual -- such as Libya's Muammar Gaddafi for example -- might do next.

Experts say analysts have in the past tried too hard to pinpoint the triggers of change, failing to recognize the greater value in identifying common behavior patterns that can more usefully predict when change is near.

"Collective behaviors are easier to understand and predict than individual behaviors," says Bar-Yam.

One example is a pile of sand growing slowly, grain by grain. A single grain dropped on the pile can occasionally cause "avalanches" of unpredictable size. The avalanches display what Bar-Yam calls "power-law behavior" -- small ones happen much more frequently than large ones.

The trigger for the uprising in Tunisia may have been a vegetable-seller publicly setting himself on fire. But the stresses in Tunisian society were building up well before that, fueled by poverty, unemployment, lack of opportunity and the rising cost of basic foods.

Bar-Yam argues the political turmoil that has brought down entrenched leaders in Tunisia and Egypt and sparked fighting in Libya are not separate from the 2008-09 financial crisis that plunged the world into recession and later began to push up food prices and unemployment rates.

"The current unrest can ... be considered as an avalanche in the sandpile of the Middle East," he wrote in a recent analysis. "Tensions ... have built up to a critical state, and the uprising in Tunisia acted as the grain which began the cascade."

The experts agree these are particularly fertile times for scientists looking for material to study change within complex systems, but say that comes as no great surprise.

An increasingly interconnected world which is using its resources almost to their limits is likely to be vulnerable to local crises that cascade and propagate.

Scheffer says experiments with smaller, more manageable complex systems such as lakes or coral reefs show that "recovery time" after small disturbances is an important factor.

"The essence is that if you perturb a system, it will be out of equilibrium for a time and then it will recover, but these recoveries will become increasingly slow as you get closer to a tipping point," he said.

As for what might happen next, the scientists warn they are in the business of predicting change, not future gazing.

"We may be able to sense which societies are close to tipping points, but that is quite a different thing from predicting what the state of those societies will be after a critical transition," Scheffer said.