Process Grammar Model

This work proposes the Process Grammar Model (PGM), a framework for describing language use along the time axis of utterance production. PGM distinguishes two modes of grammatical operation: Immediate Grammar and Adjustive Grammar.

Immediate Grammar corresponds to situations in which utterances are selected intuitively and realized promptly in real-time interaction. It does not mean that "anything can be said immediately"; rather, immediate utterances are constrained by strict and systematic rules.

Adjustive Grammar corresponds to situations in which utterances are produced through conscious judgment, revision, and structural adjustment. It is crucial for selecting appropriate words, maintaining clarity, and avoiding misunderstanding.

Importantly, Immediate Grammar and Adjustive Grammar are not separated by a fixed boundary. Their application ranges may overlap, and their relative involvement can shift depending on context, task demands, and skill acquisition. Moreover, extensive adjustment may reach a saturation point at which further revision yields little improvement.

PGM is not intended to replace earlier grammatical studies, but to provide an additional descriptive perspective that enables further development of previous approaches by explicitly modeling temporal dynamics in language use.

PGM also highlights cross-directional effects in usage (tasuki-gake effect), where the perceived effect of an expression may differ from its origin depending on how it is deployed (e.g., practiced adjusted expressions may sound polite in immediate delivery, while immediate expressions may sound natural when embedded in written or scripted contexts).

Project Main Page

• https:/github.com/yamagen/pgm/README.md

Project Main File

• https:/github.com/yamagen/pgm/pgm.pdf

Data set and Data format

• https:/github.com/yamagen/pgm/stance/README.md

License

MIT License