README

This repository contains the replication data and results for the publication

Meier-Vieracker, Simon: The dirty work of boundary maintenance. Der Topos der ‚sicheren Grenzen‘ im neurechten Diskurs. In: Bülow, Lars/Kabatnik, Susanna/Kuck, Kristin/Merten, Marie-Luis/Mroczynski, Robert (eds.): Sprachliche Grenzziehungspraktiken (= Studien zur Pragmatik). Tübingen: Narr.

The KWIC lists contain concordance lines of the CQP query [lemma="Grenze"] with a context of 25 tokens on each side. For the topic modelling as reported in the publication, the lemmatised KWIC list (without the URLs and the keyword 'Grenze') was used. The topic modelling was done with the package mallet v2.0.8 (https://mimno.github.io/Mallet/index) with the following commands:

$ .bin/mallet import-file --input kwic_grenze_lemma_url.txt --output grenze.mallet --keep-sequence --stoplist-file stoplists/de.txt
$ .bin/mallet train-topics  --input grenze.mallet --num-topics 15 --optimize-interval 20 --output-state topic-state.gz  --output-topic-keys grenze_keys.txt --output-doc-topics grenze_composition.txt

The collocations lists show the results from the collocation analyses computed with cwb-ccc Python wrapper (https://github.com/ausgerechnet/cwb-ccc) with the following settings:

$ collocates = dump.collocates(order = "log_likelihood", min_freq = 5, cut_off = 100)