Ernst Mach is often seen as a precursor of the logical positivists, an exponent of the idea that “things” are logical constructions built up out of the sensory qualities that compose the world, mere bundles of sensations. He would thus be a key example of what Graham Harman in THE QUADRUPLE OBJECT calls “overmining”. Feyerabend has shown in a number of essays that this vision of Mach’s “philosophy” (the quotation marks are necessary, according to Feyerabend “because Mach refused to be regarded as the proponent of a new “philosophy”", SCIENCE IN A FREE SOCIETY, p192) is erroneous, based on a misreading by the logical positivists that confounds his general ontology with one specific ontological hypothesis that Mach was at pains to describe as a provisional and research-relative specification of his more general proposal.
Feyerabend expounds this ontology in SCIENCE IN A FREE SOCIETY (NLB, 1978, p196-203) and in many other places, making it clear that it is one of the enduring inspirations of his work. Mach’s ontology can be summarised, according to Feyerabend, in two points:
1) the world is composed of elements an their relations
2) the nature of these elements and their relations is to be specified by empirical research
One may note a resemblance with Graham Harman’s ontology, summarised in his “brief SR/OOO tutorial“:
1. Individual entities of various different scales (not just tiny quarks and electrons) are the ultimate stuff of the cosmos.
2. These entities are never exhausted by any of their relations or even by their sum of all possible relations. Objects withdraw from relation.
The difference is illuminating. Whereas Mach leaves the nature of these elements open, allowing for the exploration of several hypotheses, Harman transcendentally reduces these possibilities to one: elements are objects (NB: this reduction of the possibilities to one, enshrined in a principle, is one of the reasons for calling Harman’s OOO an objectal reduction). Further, by allowing empirical research to specify the relations, Mach does not give himself an a priori principle of withdrawal, though here again “withdrawal” is one possibility among many. Another advantage of this ontology of unspecified elements is that it allows us to do research across disciplinary boundaries, including that between science and philosophy. Feyerabend talks of Mach’s ontology’s “disregard for distinctions between areas of research. Any method, any type of knowledge could enter the discussion of a particular problem” (p197).