Mr. Yamane has discussed the problem of grammar, meaning and logic in his two recent papers: "The Problem of Meaning (I)," Tetsugaku (Mita), vol. 50 (1967), pp. 129-149, "Sentence Structure as a Function," Tetsugaku (Mita), vol. 51, (1967) pp. 127-140. The essential points of his argument are as follows: the grammatical structure of a sentence analyzed by the transformational grammar cannot reveal its logical structure in the sense of modern symbolic logic, for this type of analysis is affected by the old-fashioned concept of subject-predicate bifurcation of a sentence. Claiming to correct such a defect, Mr. Yamane proposes a revision of the syntactic rules employing the notion of propositional function. As a result of this revision he insists that the structural meaning of a sentence will become clearer. In this paper I criticize his argument in the following way. In Chapter I, I present some fundamental points of the method of generative grammar, which will serve to make my standpoint clear. In Chapter II, I examine his rules in detail and point out several crucial faults which affect the semantic interpretation of a sentence. I conclude that his revised rules serve nothing for making the structural meaning of a sentence clearer. In Chapter III, I discuss the philosophical motivation of Mr. Yamane's proposal. He pretends to adopt R1':S→Fx, Gxy, Hxyz,…, instead of R1:S→NP+VP in the Phrase Structure Rules of Syntax. Which of these rules should we adopt? This problem cannot be solved a priori but only empirically, since the degree of success in a grammatical description of language is determined by the appropriateness of the actual data in terms of lingusitically significant generalization. He thinks that the subject-predicate analysis of a sentence is inherent to the generative grammar, and that his proposal is a fundamental revision of the entire system. But this is a misconception. Regrettably this kind of misconception is widely seen among the analytic philosophers of today.