Ellis on the limitations of dispositionalism

FIRST PARAGRAPH I have argued that dispositionalism is incompatible with the Principle of Least Action (PLA) (Katzav 2004). In ‘Katzav on the Limitations of Dispositionalism,’ Brian Ellis responds, arguing that while naïve dispositionalism is incompatible with the PLA, sophisticated dispositionalism is not. Naive dispositionalism, according to Ellis, is the view that the world is ultimately something like a conglomerate of objects and their dispositions, and that, therefore, dispositions are the ultimate ontological units that explain events. Sophisticated dispositionalism, according to Ellis, supposes that,

how things are disposed to behave depends also on what kinds of things they are, what kinds of property they have, and how these kinds of things and properties are placed in the natural kinds hierarchies to which they belong (Ellis 2005).

Further, it supposes that at the top of each hierarchy of natural kinds there is a global kind. For example, ‘[t]he global natural kind in the category of substance is that of the physical system’ (Ellis 2005). Ellis continues, claiming that the PLA is

of the essence of the global kind in the category of objects or substances. If this is so, then, of course, every continuing object must be Lagrangian, i.e. disposed to evolve in accordance with the principle of least action (Ellis 2005).

Ellis concludes that, therefore, a sophisticated dispositionalist can accommodate the PLA and its metaphysical necessity.