Modular and controlled M-grammars

Abstract

The framework described in the previous chapters is elegant and has proved its value in practice, but it also has a number of deficiencies, the most salient of which is the impossibility of assigning an explicit structure to the grammars. This may cause problems, especially in situations in which large grammars have to be written by a group of people. In the extended framework presented in this chapter, a grammar is divided into subgrammars in a linguistically motivated way and the application of rules in a subgrammar is explicitly controlled. Two kinds of rules are distinguished only one of which is involved in the isomorphy requirement. Next to that, these two kinds of grammar rules are grouped together into rule classes. The use of these divisions naturally leads to a highly modular structure for the system, which helps in controlling its complexity.¹