Description of the Parser Model
- The parser is broken up into 13 subnets.
- Input comes from the environment igniting the next orthogonal CA
when the parser asks for it.
- Output is in a case frame stored in Verb Instance, bound
to Noun Instance (and verb and noun semantics).
- When an input arrives, it turns on the appropriate semantics.
- In combination with the instances, rules are ignited and
bind words into the instances.
- The word semantics are hierarchical (from Word Net), and in
combination, they ignite the appropriate PP attachment rule
when it's appropriate.
- The overall architecture fits in with Jackendoff's tripartite theory,
and uses xbar theory.
