Tobias Rosefeldt (Berlin) gives a talk at the MCMP Colloquium (5 June, 2013) titled "Things that don't exist". Abstract: Are there things that don’t exist? Several answers seem to be possible here. You can answer ‚yes’ because you are a Mainongian and believe that existence is a discriminating property of objects, i.e. a property that some objects have and others lack. You can answer ‚no’ because you are a Quinean and believe that to exist just means to be identical to something and hence is a property of everything. Or you can be a fan of substitutional quantification and think that you can answer ‚yes’ without committing yourself to non-existing things in any ontologically interesting sense. In this paper, I want to introduce an alternative to all these views. According to this alternative, you can answer ‚yes’ to the question but nevertheless assume that (i) ‚exist’ expresses a property true of all objects and (ii) ‚there are things’ expresses objectual quantification. The reason why this is possible is that there is a literal reading of the sentence ‚There are things that don‘t exist’ in which we are using it to quantify over kinds of things and say that there are kinds that have no instances. In order to substantiate this claim, I will show that there are many cases in which natural language quantifiers such as ‚there are things’, ‚there is something’ or ‚there are Fs’ are used to quantify over kinds of things rather than individual things, and give an analysis of the intricate syntactical and semantical features of sentences in which such quantification occurs. I will then use the proposed analysis in order to show that there really is the mentioned reading of the claim that there are things that don’t exist and to explain why it has been overseen by so many philosophers. Finally, I will show how useful the insight into the linguistic structure of our talk about non-existing things is by applying it to several cases in which non-existing things are relevant in philosophy.