Realism about quantities?

Abstract

A realist about quantities holds that measurement procedures give us knowledge about quantities, where the latter are understood as entities that are in some relevant sense independent of the particular measurement procedures employed to determine them. An anti-realist about quantities holds that we can make sense of our measurement practices without introducing quantities as independent entities. With the development of formal measurement theory, which captures a wide range of measurement scales, there seems to be less need for realism about quantities. In this paper I argue in favor of a more realist approach to quantities, and lay out some conditions such a realism has to respect. A main consideration in favor of realism is the idea that in the establishment of a measurement scale, not just anything goes. A particular way of setting up such a scale is appropriate for a particular kind of quantity. I suggest that this appropriateness is best captured by assuming that the measurement aims to get right something about an independent entity. To avoid the restrictions of traditional realism about quantities, however, we must respect the idea that quantities can have different “structures”, captured by scales at different levels of fine-grainedness. I suggest that the best way to do so is structural realism about quantities. Structural realism about quantities is the idea that quantities are structures, characterized as relations among “positions”, which may or may not be occupied. Understanding quantities as structures permits us to characterize a plurality of possible values for a quantity like mass, even in the absence of a large number of massive objects. This is an advantage both over nominalistically understood representationalism, and property based realism.