Several different semantic grammars for different sub-domains can
share an underlying library of shared grammar rules, thus reusing
common underlying semantic concepts while still being developed
independently.  These can then be merged at runtime into a
comprehensive grammar for wide multi-domain coverage.  The parser
\cite{soup} applies multiple sub-grammars in parallel and stores the
outputs in a parse tree lattice.


Interlingual machine translation is
convenient when more than two languages are involved because it does not
require each language to be connected by a set of transfer rules to each
other language in each direction~\cite{NirCarTomGoo92}.  Adding a new
language that has all-ways translation with existing languages requires
only writing one analyzer that maps utterances into the interlingua and one
generator that maps interlingua representations into sentences.  The
interlingua approach also allows each partner group to implement an
analyzer and generator for its home language only.