Alonzo Church claimed that (1) 'Red' means red is properly translated by (2) 'Red' heisst rot, and not by (3) 'Rot' heisst rot, because (1) and (2) predicate the same property (meaning red) of the same object, the English word 'red', whereas (3) predicates that property of the German word 'rot'. Since (2) is plainly a contingent empirical truth knowable only a posteriori, (1) must be the same, and so too for (3). More generally, apparent semantic principles like (M) 'K' means K, (C) 'K' means L ≡ 'L' means K, or (T) 'p' is true ≡ p can be only empirical generalizations about the English language. That conclusion seems absurd while Church's reasoning seems compelling. Church goes wrong by misconceiving the grammar of displays - what the Fregean tradition miscalls quotations (because quotation marks mark displays). Displays are perceptual objects linguistically appropriated by incorporation in a syntactic structure as adjuncts of a displasionable term, a term whose extension is identifiable by presenting (e.g. by ostension) an object incorporable in an utterance. Another misconception concerns the role of symbol recurrence in the truth-securing syntax of formal truths and the significance of nonuniform substitution of synonyms in formal truths.