One objection concerns possible circularity: to identify literal use we need to know the original assignment, but how is that assignment known without first classifying linguistic uses?
The stages should be distinguished. Speakers learn established word–meaning associations through language use, instruction, and accepted conventions. Those established associations can then serve as a standard for classifying a new occurrence. The definition does not require every individual to know all assignments without evidence.
Circularity would arise if the assigned sense were inferred solely from a use whose status already presupposes that very assignment. Careful separation of the epistemic question from the definitional standard prevents the simple objection from deciding the issue prematurely.