In the last few decades a new scientific paradigm has been slowly emerging: complexity. This paradigm departs from the reductionism, determinism and materialism of classical, Newtonian science by focusing on the non-linear interactions between the components of a complex system. Out of these interactions new properties or forms of organization emerge, a phenomenon termed self-organization. The present paper will sketch the basic ideas of the complexity paradigm, and then apply them to social systems, and in particular to groups of communicating individuals who together need to agree about how to tackle some problem or how to coordinate their actions.
This is a very common situation in any kind of social interaction: individuals typically come to the table with different backgrounds, habits, ideas, cultures, perspectives and even languages. To be able to communicate at all, they should first agree about a common set of terms and what those terms mean. This is the emergence of linguistic conventions. Then they should agree about basic assumptions, such as what the situation is, what can be done about it, and what should be done about it. Finally, they will need to agree about who will do what when. If successful, this sequence of agreements will lead to a coordinated form of action, where the different members of the group contribute in an efficient way to a collective solution of whatever their problem was. This phenomenon, where a group of initially independent agents develop a collective approach to the tackling of some shared problem that is more powerful than the approach any of them might have developed individually, may be called collective intelligence.