An important step towards a scientific understanding of complex systems will be the objectification and quantification of information that carries semantics, i.e., sense and meaning. In this paper a concept is developed according to which the general aspects of semantics—such as novelty, pragmatic relevance, selectivity, complexity and others—are understood as elements of a 'semantic code". However, in contrast to its traditional usage, the term "code" does not refer to a set of rules for assignment or translation of symbols, but rather to a reservoir of value elements, from which the recipient configures a value scale for the evaluation of semantic information. A quantitative measure for the value scale is proposed.