This paper proposes a way in which the semantographic/ phonographic dichotomy recognised as fundamental in logographic (or morphosyllabic) writing systems in East Asia, the ancient Middle East, and Mesoamerica can be systematised to transcend the very different scholarly traditions in each region in order to allow valid and more meaningful cross-linguistic comparisons. A totally functional analysis with a focus on synchronic words as they occur in texts, rather than a focus on the form of signs or their etymology, ignores such formal units as the frame or even the grapheme and recognises three main compositional levels -logogram, component, and element -and the strict application of the analysis reveals cases of a fourth level, superlogogram. The application of this approach allows characterisations of writing systems that reflect the meaningful combination of signs in context and reveals greater complexity in how words are written, such as in semantic+semantic combinations, than previous analyses have recognised. It is concluded, however, that a statistical application of the analysis is prevented, not because of differences in the writing systems, but primarily because of the fundamental typological differences of the represented languages themselves.