This chapter presents standardized mean scores for individual linguistic features in the medical corpus relative to the general distributions found in the contemporary reference corpus. The present study uses multi-Dimensional analysis to show that there are systematic linguistic differences associated with micro-purpose variation within experimental research articles, but that those differences are small relative to the full range of variation among English registers. For example, academic prose, biology research articles, and Method sections in medical articles could all be considered registers, although they are defined at quite different levels of generality. In general, the more specified a register is in its non-linguistic characteristics, the less linguistic variability there will be among texts from that register. Thus, while the linguistic differences among Introduction, Methods, Results, Discussion (IMRD) sections are relatively small, these subregisters are well defined by their micro-purposes, and correspondingly the linguistic differences among the sections are systematic.