ThoughtStorms Wiki

PhilJones's academic research

notes from a short talk

Big Idea

My main interest is in understanding a number of issues grouped around the concept of the re-identifiable, persistent, individual.

Ideas associated with individuals :

  • Persistent - you can meet them at time t1, leave and meet them again at time t2.

Implying ...

  • Same body?
  • Continuity of memory?
  • Consistent with that individual at t1? (eg. can't be in two places at once)
  • Morally responsible at t2 for actions at t1.


But my question is from evolutionary psychology : Why should this concept of "same individual" evolve?

NOTE : such a question can be asked in two flavours. The historic and the asynchronous.

  • The historic question : Why, in fact, did we or species S evolve the behaviour?
  • The asynchronous question : What conditions in general encourage the evolution of the behaviour?


The method is ArtificialLife, specifically evolutionary modelling. That is, to build virtual environments where populations of agents evolve different ways of interacting. The investigation is to see what patterns we find that relate individual recognition with environmental characteristics, gross population characteristics, and other behavioural / mental characteristics.

In Practice

A simplified version of this. All my work has been done using the iterated prisoners dilemma game as the virtual environment. This is chosen as an off-the-shelf, and extremely simple model of social life.

The intuitive hypothesis : successful players of the prisoners dilemma are usually variants on the tit-for-tat strategy of doing back to your opponent what they did to you. In order to play this strategy, you need to be able to re-identify your opponent, and recall what it did to you on the previous round.

Nevertheless, looking for correlations between environmental and population characteristics is still too vast a space to explore.

I've been looking at these areas :

  • The evolutionary dynamics of the rise of Individual Recognition. That is, trying to evolve, together, improved individual recognition, and a TFT-like strategy.
  • Comparison of individual recognising strategies and kin recognising strategies. Kin recognition can lead to co-operation : co-operate with kin. Can this be the first stage on the road to evolving individual oriented strategy and individual recognition? Have a look
  • Population characteristics such as the number of breeders relative to the size of the population.
  • Environmental characteristics : playing the game in no-space (where everyone has a similar chance of meeting) and on a grid-world where players are constrained to play with neighbours.

QuestionsAboutIndividualRecognition arising from the talk


Spider's that have individual personalities? Here