Recent scholars disagree over whether Berkeley’s theory of meaning constitutes a radical departure from Locke in the direction of current philosophy of language or offers no real alternative to the semantics of Locke’s Essay concerning Human Understanding. Berkeley agrees with Locke that linguistic meaning consists in the transmission of ideas from speaker to hearer by means of words, but he does not accept the Lockean account of this transmission. Specifically he departs from Locke at two fundamental points: he insists that ideas themselves have meanings and stand in need of interpretation, and he holds that the meanings of ideas may vary with the contexts in which they occur. To accommodate Berkeley’s principle of contextual meaning the account of communication must relate not individual ideas to individual words but strings of ideas to strings of words. Words and ideas, moreover, are not isomorphic as Locke implies they are: Berkeley indicates in particular the cases of general terms and names for spiritual substances, for neither of which Corresponding ideas can be discovered. To accommodate such cases within the general theory that meaning depends upon corresponding ideas an encoding process must be introduced into the account of the verbal transmission of ideas, a process whereby verbal structures including such terms as universals and names for spirits can be related to different ideational structures in which no such terms appear. The conclusion is that Berkeley accepts from Locke the fundamental principle that meaning depends upon corresponding ideas in the mind but that he holds this relation of correspondence to be much more complex than Locke allowed: in particular Berkeley introduces structural considerations by abandoning the traditional view that words and ideas correspond on a one-to-one basis, and he requires the mind to perform certain interpretative encoding procedures in translating between verbal and ideational structures.