Whether natural language permits quantification over 'nonexistent', intentional objects is subject of a major controversy, as is the nature of such entities themselves. This paper argues that certain constructions in natural language involving 'intentional verbs' such as 'think of ', 'describe', and 'imagine' cannot be analysed compositionally without positing intentional objects, as entities strictly dependent on intentional acts. The paper also argues that intentional verbs involve a distinctive semantics, which is fundamentally different from that of intensional transitive verbs, a difference reflected in a range of quantificational phenomena.The questions whether natural language permits quantification over intentional objects as the 'nonexistent' objects of thought is the topic of major philosophical controversy, as is the status of intentional objects as such. Many philosophers deny the possibility of there being 'nonexistent' objects of thought. Others following Meinong [12], take 'nonexistent' objects of thought to be entities individuated only by a particular set of properties, and as having a weaker form of being than existence. Yes others, in the tradition of Brentano [1], admit the possibility of intentional, nonexistent objects, but take them to be dependent on an intentional act or state. This paper will argue that natural language does reflect a particular notion of intentional object and in particular that certain types of natural language constructions (generally disregarded in philosophical literature) cannot be analysed without positing intentional objects. At the same time, those intentional objects do not come for free; rather they are strictly dependent on intentional acts that generally need to be present, in one way or another, in the semantic structure of the sentence.The constructions in question display a particular dependence of intentional objects on the event argument of an intentional verb in the same sentence, a verb like F. Moltmann ( )