Markus Werning (Bochum) gives a talk at the MCMP Colloquium (11 December, 2014) titled "Making Quotation Transparent: A Compositional Analysis of an Apparently Opaque Phenomenon". Abstract: Quotation is regarded as a paradigmatically opaque context. This is due to two failures: (i) A failure of substitution: in quotations the substitution of an expression with a synonym does not leave the meaning of the embedding context unchanged. (ii) A failure of existential generalization: in quotations the replacement of a singular term with an existentially bound variable does not constitute a valid inference. From the failure of substitution it is often inferred that quotation violates the principle of compositionality, according to which the meaning of a complex term is a syntax-dependent function of the meanings of its syntactic parts. The quoted expression, so it is concluded, does not contribute its meaning to the meaning of the embedding context. The failure of existential generalization, furthermore, is taken to entail that quoted singular terms do not introduce referents into the universe of discourse and are hence referentially vacuous. Since compositionality and referentiality are two constitutive principles of semantics, many authors view quotation as an extra-semantic and hence mainly pragmatic phenomenon. The purpose of this paper is to turn the tables and to re-establish the genuinely semantic character of quotation. In the first step I recall that in natural language many phenomena indicate that quoted expressions, at least sometimes, do contribute their meanings to the meaning of the context. This is evident in the case of mixed quotation, but arguable also the case in direct quotation and in many other cases. Moreover, it is argued that the referents of quoted singular terms can often be anaphorically referred to, not only in mixed and direct quotation, but even in some cases of pure quotation. In the second step the inference from opacity to non-compositionality and non-referentiality is reviewed with greater scrutiny. It turns out that additional premises have to be presupposed to make this inference go through. In the third step the pivotal premise that a quoted expression corresponds to a uniform syntactic part of the quotation will be rejected and replaced by the assumption that the quoted expression contributes a syntactically complex structure, which primarily contains phonological and other sub-symbolic information. However, due to a particular semantic interpretation of the quotation marks, the standard meaning and referent of the expression is recovered such that the quoted expression can contribute its standard meaning and referent to the discourse in many circumstances.