corpuslinguïstisch

Corpuslinguïstisch, or corpus linguistics, is a branch of linguistics that studies language by collecting and analyzing large electronic collections of texts called corpora. It combines quantitative methods, such as frequency counts and statistical analyses, with qualitative analysis of patterns in actual language use. The central idea is that language phenomena are best understood by examining large, authentic samples rather than isolated examples.

Corpora vary in size and scope and can be general, covering broad language use across genres, or specialized, such as spoken dialogue, academic prose, legal texts, or learner language. Notable benchmark corpora include the Brown Corpus, the British National Corpus, and the Corpus of Contemporary American English, with many language-specific and domain-specific corpora, as well as web-derived and parallel corpora for translation studies. Corpora are typically annotated for features such as part of speech, lemmas, morphology, syntax, and semantics, enabling advanced queries and analyses.

Common methods include concordance analysis, keyword and keyness analysis, collocation and dispersion studies, frequency profiling, and distributional semantics. Tools range from corpus query systems to tagging and parsing pipelines and scripting libraries. Applications span lexicography, language teaching, translation studies, sociolinguistics, and natural language processing resource development.

Limitations include representativeness, sampling bias, annotation quality, and the resource demands of less-studied languages. Ethical considerations involve data licensing and privacy of spoken corpora. The field continues to evolve with web and multimodal corpora, annotation schemes, and practices that emphasize reproducibility.