Scott talked about Equilogical Spaces.
Trying to think about 3-ary trees with
an equivalence relation on their leaves
where beta reduction is defined
(t1, t2, t3) -> [t1/t2]t3
(substitution is sort of pattern matching,
where a 'match' occurs up to equivalence
of leaves)
I wonder if there are any nonnormalizing trees?
