Talk:Normalisation by evaluation

False claim about normal forms?

I think this is false:

By induction on the structure of types, it follows that if the semantic object S denotes a well-typed term s of type τ, then reifying the object (i.e., ↓^τ S) produces the β-normal η-long form of s

e.g. SYN (app (lam "x" (var "x")) t) where t is some term of basic type. This denotes a well-typed term, yet reifying it does not produce the normal form. It's only after applying meaning that we get a normal form. Can someone confirm? U25506 (talk) 15:55, 10 May 2011 (UTC)

I think "S denotes s" is equivalent to "the meaning of s is S" (and SYN is not a "real" semantic object except for values of base types).Ejrsmi (talk) 16:47, 10 January 2013 (UTC)