TM and NLP

Greetings:
I have gradually understood the relationship between TopicMap 
standards, and the notion of a semantic network - focusing on 
meanings (semantic) by treating any linguistic/non-linguistic string 
as a "topic" subject to the node-association terminology.  But I'm 
still stuck when I try to figure out the relationship between NLP 
databases, which focus on morphological/word/phrasal units of 
language, and associations among them (e.g., parsing and word/phrase 
building).

Has anyone seen an application of topic maps to linguistic data?  I'd 
appreciate comments, suggestions, pointers.
Thanks,
Bob Parks