McKitrick, J. A case for extrinsic dispositions. Australasian Journal of Philosophy, 81(2), 155-174. doi:10.1080/713659629
[Abstract]Many philosophers think that dispositions are necessarily intrinsic. However, there are no good positive arguments for this view. Furthermore, many properties (such as weight, visibility, and vulnerability) are dispositional but are not necessarily shared by perfect duplicates. So, some dispositions are extrinsic. I consider three main objections to the possibility of extrinsic dispositions: the Objection from Relationally Specified Properties, the Objection from Underlying Intrinsic Properties, and the Objection from Natural Properties. These objections ultimately fail.
[Citing Place (1999b)]  
Citing Place (1999b) in context (citations start with an asterisk *):
Section I. Marks of Dispositionality
* [...] a disposition has a characteristic manifestation. Some event-type is associated with a disposition that occurs when the disposition is “triggered.” The manifestation of fragility is breaking. The manifestation of solubility is dissolution. The manifestation need not occur for something to possess the disposition. The fragile glass might never shatter. Fn 3: The idea that dispositions have manifestations is standard among philosophers, despite widely divergent views on the nature of dispositions. For example, Ryle says “The tendency to ruminate and the habit of cigarette smoking could not exist, unless there were such processes or episodes as ruminating and smoking cigarettes” [1949: 117]. Armstrong says “we distinguish between the thing’s disposition and the manifestation of that disposition; between the brittleness of a piece of glass and its actually breaking. We recognize further that having the disposition does not entail manifestation of the disposition: a piece of glass may be brittle and yet never break” [1973: 8]. U. T. Place says “we invariably characterize a disposition in terms of its manifestations” [1999: 226].