Sufiksitest

Sufiksitest is a methodological framework in linguistics and natural language processing for evaluating how suffixes attach to base forms and how productive those suffixes are across languages with suffixal morphology. It combines experimental tasks, corpus data, and computational modeling to assess acceptance, processing costs, and the likelihood of forming new words with given suffixes.

The term is typically used to study morphological productivity and rule-governed affixation. It can be applied to natural languages with rich derivational or inflectional suffixes as well as to constructed languages or language technologies that rely on suffix-based grammar. The approach emphasizes both descriptive evidence (how speakers actually use suffixes) and prescriptive or model-based evidence (how well automatic systems can parse or generate suffixal forms).

Methodologically, sufiksitest involves several steps. Researchers select a base lexeme, design a set of nonce or real suffixes, and generate novel words by attaching suffixes to the base. They then present stimuli to native speakers or language models, collecting acceptability judgments and processing measures such as reaction times or eye-tracking data. The data are analyzed with statistical models to estimate suffix productivity, distributional patterns, and processing costs, often comparing real versus nonce suffixes.

Applications extend to linguistic description, language education, and technology development. In NLP, sufiksitest informs morphological analyzers, spell-checkers, and language models for agglutinative or highly inflected languages. It also aids lexicography and language pedagogy by clarifying which suffixes are likely to be productive or transparent to learners.

A simple illustrative example uses a base form and several nonce suffixes to gauge acceptability, helping to map the boundaries of suffixal productivity in a given linguistic context.