The aim of the present thesis is to introduce the topic of dummy or grammatical subjects in English, Norwegian and German and to investigate their relative frequency in a multiple parallelcorpus set up by the University of Oslo. The mainresearch areas behind this investigation concern contrastive languagetypology of individual syntactic, lexical and semantic devicesemployed in information processing on the sentence level, aswell as discourse structuring on the text level. Thestudy is biased towards English, as all corpus data are takenfrom the English-Norwegian-German part of the Oslo Multilingual Corpus. The thesis briefly explores such syntactic patternsas presentatives, clefts, impersonal passives, specifyingconstructions, extraposition, and other constructions that makeuse of semantically empty – dummy - subjects. Through the comparison of the individual dummy subject construction types and their translation equivalents, I will attempt to illustrate that individual subject choices often reveal interesting similarities, aswell as differences, between the investigated languages. Thesyntactic patterns under the study are often interrelated on different levels and provide an interesting background for the discussion of the grammatical possibilities of the individual languages, their information structural resources in following the functional sentence perspective, and their preferences in the assignment of thematic roles to individual sentence subjects. The qualitative analysis presented in this study treats the level of syntax, lexis and discourse semantics as joint premises.