This paper first draws a distinction between English inanimate sentences (EIS) and English sentences with inanimate subjects (ESWIS) and then, with a written text and its corresponding oral discourse from M. A. K. Halliday as research data, makes a contrastive case study of the distribution and pragmatic features of EIS and ESWIS in the text and discourse. The study shows: 1) EIS are evenly distributed in both the text and the discourse, while ESWIS mostly exist in the text; 2) EIS in the written text appear overwhelmingly in the passive construction and EIS in the oral discourse occur mostly in the S+V+C pattern, while ESWIS in both the text and the discourse exist chiefly in the S+V+O pattern; 3) ESWIS exist abundantly in the written text and appear scarcely in the oral discourse with different pragmatic features: ESWIS in the written text enjoy a high degree of animacy and metaphoricity and are therefore characterized by their vividness, while ESWIS in the oral discourse bear a low degree of animacy and metaphoricity and are therefore typical of their plainness.