In this paper, two methods appropriate for the visual and quantitative inspection of multivariate data are applied to a corpus of narrative texts in order to investigate the behavior of the Tok Pisin suffix -im. The distribution of the suffix is subject to considerable variability. While this problem has been noted in the literature and various solutions have been proposed, there are a number of verbs whose behavior is not easily accounted for. The methods used are complementary explorative tools for the analyst. They are deployed in an effort to ascertain which factors govern the encountered patterns, and what weight these factors carry relative to each other.