It matters, therefore, above all that falsification remains possible, that it is not prevented by dogma or even the vested interests of the scientific community. We cannot be sure what the good society is like, we can only advance projects to this end. Such projects may turn out to be unacceptable or inappropriate; in fact, debating their pros and cons is what life in the open society means. It matters that change remains possible, that it is not prevented by tyranny or cartels.
Popper was right when he pointed to profound differences between the natural and the social sciences. Time, history rather, signified the key difference. If Einstein proves Newton wrong, Newton was always wrong. If a neosocial-democratic project replaces a neoliberal one - Clinton after Reagan and Bush, Blair after Thatcher and Major - this may mean that a project which was right at its time has come to be regarded as wrong. Perhaps it even means that all projects will in due course be "wrong"; history knows no "truth".
Society not only has history but it is also heterogeneous. It must remain possible to remove governments without force, no less, but no more. Applied to economic processes, the notion that comes to mind above all is that of the market. It alone leaves open the door to changing tastes and preferences, as well as to new "forces of production".
Schumpeter's world of "creative destruction" by entrepreneurs is in some ways the economic version of progress by falsification. In society in the more diffuse sense it is harder to find the equivalent.
Probably the notion of pluralism is relevant here.
The notions of democracy, the market economy, and civil society must not mislead anyone into believing that there is only one institutional form to give them reality. All that remains essential for open societies is that there are rules of engagement which allow the continuation of the process of trial and error. The whole point of the open society is that there is not just one way, nor are there two, or even three, but 101 - that is, an indefinite, unknown and unknowable but large number.
This is why 1945 and 1989 are such critical and exhilarating dates in the dismal history of the 20th century. They mark the defeat of the enemies of the open society, and thus not the victory of another "system" but the opening up of wide and varied horizons of opportunity.
Diversity and variety are now the name of the game. There are many ways to try, and it is wrong to recommend, let alone to impose, just one version of democracy or of market economy to those liberated for the new openness. There are, on the other hand, new dangers also, and they have to do with ambiguities and problems arising from the concept of the open society.
If trial and error define the open society as well as the advancement of knowledge - what if people stop trying? What if no one endeavours to discover anything new?
Like the dramatic moments of lost theories and lost elections - perhaps even of the "crisis of global capitalism" - normal scientific, political and economic activity in the open society has to be public. Trial and error are public events, visible, open to critical appraisal, an invitation to all to participate.
This, rather than some capitalist crisis, is in my view the deepest problem of some south-east Asian countries. This is also where the postcommunist countries of East and East-central Europe find it hardest to move towards open horizons.
If the open society is not also an active society, it will soon cease to be open. But openness by itself does not create the activity which is needed. What does? I have long liked Immanuel Kant's strictures against the Arcadian longings of many, and his insistence that our "unsociable sociability", our penchant for power and possession and the resulting social conflicts, were the foundation of our freedom. Popper has certainly set an example, not just by his ideas but by his often fierce polemics, and by indomitable curiosity.
Critical discourse and political conflict are the lifeblood of a world in which people never cease trying, erring, and trying again.Reuse content