corpussuch

Corpussuch is a neologism in linguistics and digital humanities describing a framework for building and analyzing language corpora in ways that emphasize representativeness and comparability across domains. The term is used to refer to practices that document corpus compilation, sampling, annotation, and analysis to ensure that findings generalize beyond a single data source.

Origin and concept: The name combines corpus and such, signaling the focus on constructing corpora that exemplify "such" language use across contexts. It is not a formal standard, but a conceptual guide used in scholarly discussions and some research projects.

Principles and methods: Emphasizes transparent methods: specify data sources, sampling strategies (random, stratified), annotation schemas, versioning, and evaluation metrics. Encourages using multiple corpora to test generalizability, and reporting limitations.

Applications and examples: Used to study lexical variation, pragmatics, and discourse patterns across genres (news, social media, academic texts). In computational linguistics, corpussuch-informed corpora support model training and bias analysis.

Reception and critique: As a concept, it promotes reproducibility but may be resource-intensive and risk over-emphasizing representativeness if sources are biased. Critics call for careful privacy considerations and inclusive sampling.

See also and references: See also Corpus linguistics; Reproducibility; Data ethics.