Frege's Error? Compositionality, Inference, and Final Causes
Daniel Everett
July 2024 

This paper argues that compositionality is a form of inference that needs to be embedded into a larger theory of inference, e.g. Peirce's triadic theory (abduction, induction, deduction). Otherwise Fregean compositionality is not only insufficient but misleading, creating artificial problems (e.g. Pseudogapping, Antecedent-Contained Deletion, Sluicing and other pseudoissues). The paper also makes a case that the fixation on propositions/sentences as the beginning point of linguistic analyses misses significant generalizations about human languages. It is also argued here that Peircean semeiotically-based inference theory and Peirce's Existential Graphs provide much simpler and more useful theories of meaning-form correspondence, by analysing linguistic forms as "anchors" for inference and restricting inferential space. The paper also claims that the final causes (in the Aristotelian sense) relevant to modern linguistics are sentence structures and meanings (for Frege and Chomsky), unlearnability (Chomsky), and the logic of discourse (Peirce).

Format:	[ pdf ]
Reference:	lingbuzz/008270 (please use that when you cite this article)
Published in:	work in progress
keywords:	sluicing, pseudogapping, antecendent-contained deletion, inference, abduction, induction, deduction, compositionality, frege, peirce, transparency, recursion, existential graphs, vp-ellipsis, syntactic islands, semantics, pragmatics, grammar types, semantics, syntax
previous versions:	v4 [July 2024] v3 [July 2024] v2 [July 2024] v1 [July 2024]
Downloaded:	582 times

 

[ edit this article | back to article list ]