Corpusranging

Corpusranging is a term used in corpus linguistics to describe the systematic exploration of language corpora to map the usage range of a linguistic item across contexts. It involves examining how a word, phrase, or construction appears in different genres, domains, time periods, and demographic groups, indicating where and how often it occurs.

The aim is to establish bounds of usage and variation, providing data for lexicography, language teaching, and natural language processing. By identifying typical contexts and outliers, corpusranging supports decisions about sense distinctions, register, and domain adaptation.

Researchers compile or select multiple corpora or subcorpora, stratify data by variables such as genre and time, and compute statistics including frequency, dispersion, and collocation profiles. They may track diachronic change, visualize ranges, and apply statistical tests to compare usage across groups.

In lexicography, corpusranging informs sense inventories and usage notes; in NLP, it guides domain adaptation, terminology extraction, and error analysis; in sociolinguistics, it enables studies of variation and change.

Findings depend on corpus representativeness, annotation quality, and sampling strategy. The identified ranges reflect observed data rather than universal linguistic constraints, and results can be sensitive to methodological choices.

See also: corpus linguistics, frequency analysis, dispersion, collocation analysis.