Complexity and Social Networks Blog: John Casti "Why the Future Happens: Socionomics and the Science of Surprise"

February 2008
Sun	Mon	Tue	Wed	Thu	Fri	Sat
					1	2
3	4	5	6	7	8	9
10	11	12	13	14	15	16
17	18	19	20	21	22	23
24	25	26	27	28	29

« Yochai Benkler on "The Wealth of Networks" at Havard Law School | Main | Sampling Strategy Online Forum »

21 April 2006

John Casti "Why the Future Happens: Socionomics and the Science of Surprise"

Here is the abstract of John Casti's talk, from today "Why the Future Happens: Socionomics and the Science of Surprise"

Please feel free to comment.

This talk presents the implications of recent work on wave-like patterns in social phenomena for characterizing and predicting the flow of human events and actions, such as the outcome of political elections, trends in films and fashion, the outbreak of war, and the rise and fall of civilizations. The talk will show that all such collective human social events are generated by the changing social mood in a population, and that the changes in this mood follow patterns that are predictable. This fact has led to the emerging field of socionomics, which is nothing less than a "science of surprise". The story of the development of socionomics and its mode of forecasting social trends will be presented through numerous examples and stories, as well as illustrations of how socionomics provides a systematic, coherent tool for predicting changing trends in the overall social mood. The talk shows that these trends exist on all time-scales---minutes to decades---and can be measured by the gyrations of financial market indexes, such as the Dow Jones Industrial Average. Thus, the talk illustrates how to actually predict "surprises", the turning points in social trends. So in a very real sense socionomics provides what amounts to a telescope for seeing the shape of the future. The "take-home" message from this talk is twofold: The future is predictable in exactly the same probabilistic way that the weather is predictable, and that thoughts cause action and events---not vice-versa! Individual human thoughts are gathered together through the herding instinct hard-wired into every mammalian brain.These individual moods are like "bets" people place about the future. The bets then self-organize into an overall collective social mood, which after an appropriate period of time depending on the nature of the event, gives rise to things like wars, election results and styles in popular culture. This line of argument is exactly the opposite of that usually put forth by academic thinkers, Op-Ed writers, intellectual commentators, and other assorted pundits in their attempts to "explain" the flow of human events. The conventional explanations mostly go from "actions -- moods -- thoughts" instead proceeding in the "socionomic direction", which puts things exactly the other way around.

Posted by David Lazer at April 21, 2006 12:43 AM

Comments

Here are a few more comments and questions on John Casti's talk.

1.) As John mentioned this is only an early proposal, so the empirical side of this is somewhat casual. However, an interesting question is in the end, when this gets more formal, how will this be tested? At this point you will need to take a stand on what you think these mood measures are predicting, and move away from a set of favorable examples. You don't want to be accused of picking and choosing the cases where this works.

Two complexity related questions:

2.) You continued to downplay any kind of feedback from the system and events into the "social mood". This seems like a very un complex systems approach, and also not easy to justify. Why isn't mood itself dependent on the actions and events of the system and agents (policies, institutions, and all things dependent on today's mood).

3.) Another question is what is the "aggregate mood" and could dynamics depend on the distributions of individuals moods? I hinted at this in the discussion, and you got a quesiton on this. Most of the agent-based simulations I do are based on the dynamics of a population distibution through time. This is why I often have to turn to the computer since these objects are difficult to analyze. To some people in the agent-based world, aggregating mood could be destroying a lot (if not all) of the interesting information in the system.

Posted by: Blake LeBaron at April 21, 2006 10:29 PM

My understanding was that one of the main ideas of Casti's hypothesis is that social mood causes social events, rather than vice-versa. When the social mood is positive, positive things will happen, and when the social mood is negative, negative things will happen. This perspective is contrary to popular conceptualizations where the mood might result from events.

These two distinct perspectives may very well be testable with the kind of data Casti presented.

If events cause mood, then when a negative mood trend transitions to a positive mood trend, we would expect to see many positive events around the nadir that served to turn things around. Similarly, transitions from positive to negative should have indicative negative events at the peak.

If Casti's hypothesis is accurate, then we would expect few to none positive events at local minima of the mood curve, and few to none negative events at local mood maxima.

I think these transitions are important loci for evaluating the tenability of Casti's hypothesis.

Posted by: Brian Rubineau at April 23, 2006 4:09 AM

Complexity and Social Networks Blog

February 2008

4

5

6

8

9

19

24

25

26

28

Editor Login

Convener in chief:

Editors:

Categories

Archives

Recent Entries

Recent Comments

Notification

John Casti "Why the Future Happens: Socionomics and the Science of Surprise"

Comments

Notification

Post a comment