The data were used in a corpus-based study that investigates the variation of concessive constructions across nine varieties of English. Concessive constructions are here taken to consist of a subordinate clause linked to a matrix clause using one of the three subordinating conjunctions 'although', 'though' or 'even though'. For each occurrence, the data contain information concerning its semantic properties, the position of the subordinate clause, the conjunction that was used, the finite or nonfinite status of the subordinate clause as well as its length. Further, each token is annotated for variety, mode of production (spoken vs. written) and genre (or text type). It is also possible to model the text frequencies of conjunctions and semantic subtypes, since in the respective data tables counts are given for each text in the corpora, along with the total word count per text.
AntConc, 3.5.7
R, 3.6.2
RStudio, 1.2.5033
R package 'brms', 3.6.2
R package 'rstan', 2.19.3
R package 'StanHeaders', 2.21.0-1
R package 'lattice', 0.20-38
R-package 'latticeExtra', 0.6-29
https://www.ice-corpora.uzh.ch/en.html [ICE-website at the University of Zurich]
Kirk, John & Gerald Nelson. 2018. The International Corpus of English project: A progress report. World Englishes 34(4): 697–716. https://doi.org/10.1111/weng.12350
