This paper is an attempt to reinvestigate speech act classification ― especially the taxonomy of illocutionary acts ― that has been a point in dispute among many thinkers, and to apply macroscopic dimensions (speaker-hearer or participants, state of affairs of the world, and time dimensions), which I think have been almost ignored, to it. Many philosophers and linguists have thought that in classifying illocutionary acts different dimensions are required. This is what I would call the microscopic classification. But it is difficult to pinpoint where such different dimensions begin and where they end. In classifying speech acts, therefore, we might need hundreds of thousands of dimensions and even more. If the primitive illocutionary force indicators (Transformational, Acknowledgment and Informational force indicators), which are closely tied to three primary dimensions, are adopted as the fundamental notion for classifying speech acts, an indefinite number of speech act types can be converged into a small number of basic categories. The theoretical account of speech act classification will pave the way for the actual analysis of our ordinary, everyday speech activities.