Generic reference has mainly been studied by philosophers and logicians, and their interest has been on questions of truth conditions of genericness (as opposed to universal quantification) and their formal representation. There has only been little interest in distinguishing types of genericness as reflected by linguistic expressions. Also most linguists working on generic reference have not been concerned about distinguishing between different types. For cognitive linguists generic reference should be a highly interesting field of study for the following reasons. English and, for that matter, probably most other languages, do not have exclusively generic forms of reference. Rather, the referring expressions used for generic reference, i.e. for reference to classes, are the same as those used for non-generic, or individuative, reference. Referring expressions are thus systematically polysemous. The senses of polysemous expressions are related and, as cognitive-linguistic studies on polysemy have shown, the links between the senses are motivated. Also the relationship between individuative and generic reference should be expected to be motivated. Obviously, reference to individuative objects is more basic than reference to generic classes. It is, therefore, to be expected that the meanings of individuative reference are extended to generic reference in a motivated fashion. This paper wants to show how types of generic reference are motivated from their corresponding types of individuative reference. (Günter Radden)