The ‘privileged’ situation of English-speaking world (Halliday 1994: xxx) has made English the most studied language in the world. Indeed, it is no doubt that the existence of alternative versions of SFG for English demonstrates the development of Systemic Functional tradition in depth both qualitatively and quantitatively. In fact, on entering Phase 5 of the COMMUNAL Project in 2002, as developed by Fawcett and other scholars, Fawcett (2002) makes it clear that ‘this already very large generative lexicogrammar has the potential to be […] the largest fully integrated, operational model of any language in the world’. Along this line, it may be possible to say that researches on Japanese has the potential to be invaluable in its contributions to SFL in the global trend. The aim of this presentation is to make an attempt to show the way, in broad terms, to advancing step by step from one stage to the next in developing the Japanese version of Systemic Functional Grammar, which I refer to specifically as a ‘proper grammar’.To do this, the present study will be to pinpoint the major problems that we currently face in describing Japanese grammar in terms of SFL. Some are derived from applying the method of ‘transfer comparison’ to the description of Japanese grammar, which may cause certain difficulties in developing a 'homegrown' framework for analyzing natural text (as Eggins and Slade 1997 does for English). Yet, more importantly, other problems reside in the lack of the integral model of human communication that handles the whole processes of meaning-making in a specific language.