Developing comprehensive criteria of adequacy
Butler (2009a, b) argues that an adequate model of the language system shouldaccommodate cognitive, sociocultural, discoursal, acquisitional, typologicaland diachronic dimensions, and observational evidence from corpora, experimentsand intuition. This paper asks if such reconciliation is possible. It arguesthat language is composed of accreted subsystems that render the linguisticsystem inherently complex in each dimension. This hybridity explains the difficultyin constructing Butler’s macro-model, but also indicates how it mightbe done. Subsystems that add complexity in one dimension are often explainedby another, e.g. sub-patterns for English plural formation arose for socioculturalreasons (Classical borrowing); typological exception groups (e.g. DirectorGeneral) have a diachronic explanation. Thus, future modelling will benefitfrom the flexibility to cross-refer between dimensions.