If you are looking at this paper as one of a sampling of WPE II papers in order to get a sense of of what kind of paper to do for your WPE II paper, please beware.

The topic of this paper is unreasonably broad for the scope of the WPE II exam.

In selecting this topic I expected that the modelling of action in different domains would reveal that there was a common subset of the components of action (perhaps a set of primitives) that would have to be represented by any system that needed to represent the physical movement of agentive beings in the world.

That expectation was naive, at best, and was not satisfied.

While comparing the approaches of the three papers was very interesting and I learned a lot, from the WPE II exam paper point of view it was frustrating. The reason for this is that each of the three papers I reviewed came from a different branch of the AI "family tree". Thus each paper had an extensive history behind it, and it was not possible in the bounded time of the exam to do justice to the literature for each of the domains: robotics and animation/avatars and logical representations.

So my advice to me next time would be to pick a narrower topic.