Fw: empty verbs

[Robert Levine]

A trace meets a single description, no?---one where the LOC and SLASH
set membership are token-identical and the PHON is null. By the same token,
the transitive verb `kicks' meets a single description. The fact that
there are an unboundedly large number of signs that satisfy it doesn't
mean that there isn't a single verb `kicks'. If you want to argue that
there are a vast number of traces, then you are in effect arguing that
there are as many verbs `kicks' as meet a lexical description which includes
the specification [COMPS <NP>]...

Bob