The concept of a trope (understood as an individual property and not as a figure of speech) plays an important role in contemporary analytical metaphysics. It is, however, often far from clear what the logic of this concept really is. Indeed, there are two equally important intuitions underlying the concept of trope, two intuitions that generate two quite different conceptual frameworks. According to the first intuition, a trope is a particularised property – a property taken as an individual aspect of (...) a particular object. A trope is thus the colour of this table or the shape of this sculpture. In the light of this conception tropes are, as it were, direct abstractions from concrete individuals. According to the second intuition a trope is still such an individual (and abstract) colour or such an individual (and abstract) shape, but now it is not only abstracted from the concrete individual of which it is a trope but also taken in abstraction from the very fact that it has been abstracted this way. We can imagine that the cognitive access to these tropes is obtained by a kind of a double abstraction. All this can sound puzzling at first blush, but in what follows I hope to make the matter a little clearer. Let us anticipate briefly. On the second view, tropes are construed not so much as abstractions from concrete individuals, but rather as primitive items of which the concrete individuals are composed. Tropes so construed are, as it were, metaphysically prior to concrete individuals. We will call tropes of this kind unstructured tropes, whereas tropes conceived as aspects of individual things will be called propositional tropes. This terminology is justified because tropes understood as aspects of things seem to involve something like propositional structure, a structure that tropes conceived as primitive “building blocks” of reality definitely lack. It is sometimes believed that these two concepts of trope are nearly equivalent, that the difference between them is for the most part verbal. Yet it will turn out that there is an important ontological difference here. Propositional tropes are non-trivially propositional and unstructured tropes are non-trivially deprived of such a structure (althought not necessarily deprived of any dependence-links with respect to other tropes). It will be argued that tropes are metaphysically interesting only on the second conception (i.e., as primitive, unstructured items), whereas only on the first conception (tropes as propositionally structured entities) can they function as semantically efficient truthmakers. The moral of the paper will be that our trust in tropes must be mitigated. A trope ontology could plausibly be viewed as a serious metaphysical alternative, but its semantical possibilities should not be exaggerated. (shrink)