In this paper, we report the work on developing a multilingual digital linguistic database that aims to provide overall information a linguistic item carries in a language, and its cross-linguistic morphemic equivalent in other languages. It is conceptualised as a model of human knowledge of a language, and its description and architecture is an effort towards modeling such linguistic knowledge. From computational and programming aspects it throws an enormous challenge as it has many-to-many relations across languages, scripts, orthography, fields and entries. To accomplish such linkages among languages and between different types and kinds of information related to the linguistic item, an idea of a ‘Concept Set Model’ is discussed.