Abstract
The paper deals with (meta)linguistic categorization and its role in linguistic/grammatical research. Following the presentation of the starting point on how to assess distinct categorizations, it is shown how the absence or presence of theoretical positions on categorization in languages (structuralism, generativism, usage-based approaches) affects linguistic descriptions, and examples of inconsistencies are also presented.
Subsequently, it is suggested a new categorization perspective can be used to overcome existing categorizations. The paper also presents the position that the study of language concerns causal processes applied in multiple frames and concludes that doing distinct and compatible categorizations can truly serve as a spiritus agens of new findings in linguistics.