Contextualism probably should not be transformed into empirical realism (an oxymoron). Words ought to be used for pragmatic purposes rather than for theoretical juxtapositions and relationships as if they were mathematical entities of Platonic realist states. Skepticism is great. Everyone in the 20th century offered refutations of Descartes it seemed, for one reason or another. Descartes’ method was a correct tool for the day to get people thinking about the nature of things and it worked quite well. It still is useful.
Skepticism vs Certainty in language and the correspondence to word and objects. With Pierce one might use language trielectically to attempt to discover new things and it potentially could be productive; at least for an A.I. constructing with tremendous power a vast number of theoretical propositions about the unknown.
I have been influenced quite a lot by the linguistic philosophy of W.V.O. Quine, Saul Kripke, P.F. Strawson and others. Word and Object by Quine and Individuals by Strawson consider subjective epistemology vs empiricism. Quine wrote The Two Dogmas of Empiricism that invalidated the empiricist outlook convincingly. The boundary between intentional and extensional thought and words are never concisely delineated as empiricists prefer. Words and knowledge about ‘external objects’ are made from thought too. One cannot just speak about external objects as if were not also defined with the mind. The Heisenberg uncertainty principle regarding the observer effect on quantum states is consistent with the ambiguous linguistic paradigm for epistemology.
Quine’s ‘Ontological Relativity and Other Essays‘ describe language lexicon-Universes existing each within their own ontology. Each lexicon or ontology is self-standing and the terms are not imbued with an element of Platonic realism about them. To a certain extent words and word phrases in some epistemological debates chase around lexical meanings and agreements. A goldfinch or any other perceptible bird-as-an-object is what it is regardless of what it is perceived One may use Sartre’s terms well here about objects in-themselves, for-themselves and for-others. Two philosophers perceiving an elephant each might believe it either a goldfinch or a woodpecker variously, and be wrong about the object in-itself though in agreement that they are correct..
Language truth values in corresponding words to objects have functional accuracy value rather than truth for-itself. A truth theory such as correspondence theory wherein accurate and commonly associated word relations to objects determine the truth of propositions about actual objects perceived, has pragmatic communication value instead of transcendent accuracy. Who wants to name an iceberg the Rub al Khali desert? That is only right considering the paradigm of humanity existing within a quantum Higgs field wherein Adam was given to assign names to everything for-himself.