To What Extent is a Group an Individual? (Rohit Parikh)

The Logic and Metaphysics Workshop will meet on May 14th from 4:15-6:15 in room 3309 of the CUNY Graduate Center for a talk by Rohit Parikh (CUNY).

Title: To What Extent is a Group an Individual?

Abstract: Dennett in his Darwin’s Dangerous Idea (1995) and Kinds of Minds (1996) discusses an evolutionary hierarchy of intellectual progress. He calls the hierarchy the ‘Tower of Generate-and-Test,’ where there are five kinds of creatures.  These range from  ‘Darwinian creatures,’ organisms which are blindly generated and field-tested, to Popperian creatures which can make plans,  to creatures like human beings who use ‘language’ to communicate with others like them. One could ask, “at what level, if any, do groups belong” if indeed we can regard them as individuals or as intentional beings?  Since they do use language, one would think, they are creatures of this last level.  But difficulties arise in thinking of groups as even Popperian. In order for a group to have a real identity, it needs coherence in its “views” and in its actions.  To think of it as a game theoretic opponent (or partner) one needs a certain amount of predictability. Such predictablility is not always absent.  We know quite well how “Russia,” thought of as an agent, will respond in case of a nuclear attack.  But “the Republican party” or “Republicans in Congress” might be less predictable in their response to say the election of Conor Lamb. Two kinds of theoretical issues thus arise. One is epistemic coherence which can exist only if the group possesses mechanisms for intra-group communication.   An army preparing to wage a battle needs scouts to gather information and to transmit it to troops.  A university needs an internal email system. The other is exhibiting coherence of views where issues like the Arrow theorem or the judgment aggregation paradox may arise.  A group where power is more concentrated, at the extreme in one individual, is likely to be more predictable and more consistent in its response.   It applies less to a more diverse group like the Democratic party. So a better question to ask than “do groups exist?” is “to what extent is a given set of individuals (with a name) an individual, and on what issues?”  In other words we suggest an algorithmic and game theoretic alternative to the ontological question.  We will offer some answers while avoiding a surfeit of mathematics.​