spellingspisky

Spellingspisky is a term used in digital linguistics to describe the study and documentation of spelling variation across languages and historical periods, with a focus on nonstandard and creative spellings found in everyday writing. It can refer to both a methodological approach to collecting spelling variants and a repository that stores those data for analysis.

The name is a portmanteau of "spelling" and "spiky," intended to evoke the jagged, irregular patterns often visible in informal text, user-generated content, OCR outputs, and other nonstandard orthographies.

Scope and methods: Spellingspisky projects typically center on English but may include other languages. Data sources include digitized books, forums, social media, and OCR corpora. Records capture the base form, observed variant, language or dialect, date, source, and annotation confidence. Analyses combine crowdsourced annotations with automatic phonological alignment, normalization, and frequency analysis.

Applications: The collected data support improvements in spell-checkers and natural language processing, and they aid historical linguistics and education by illustrating how orthography reflects phonology, dialect, and technology.

History: The concept began circulating in online linguistics communities in the early 2020s. Pilot projects published datasets and methods, and several universities formed collaborations to create open repositories intended for researchers and educators.

Criticism and limitations: Observers point to potential data biases, copyright and privacy concerns, annotation reliability, and difficulties in cross-dialect comparison. Proponents emphasize transparent documentation, available licensing, and robust methodological controls.

See also: orthography, spelling variation, corpus linguistics, natural language processing, lexicography.