with reality—intersubjective agreement; method: rational argumentation—rhetorical persuasion; historical development: progress—changing styles; core doctrines: truth—tradition.
In this paper we examine one background issue, the language of classical physics. 2Though my primary concern is the relation between classical and quantum physics, rather than the science wars, I will indicate the bearing this has on the issues just schematized. The language of classical physics is a linguistic parasite. It evolved from an Indo-European core but can now attach itself to Japanese, Arabic, or any other language while preserving its basic structure and inducing mutations in its successive hosts. It is a language of discourse, particularly the discourse between experimenters and theoreticians. We are particularly concerned with its categorial structure, or the cluster of core concepts. Two considerations motivate this special focus. First, for familiar reasons, the categorial structure of a spoken language reflects and shapes the ontology, or descriptive metaphysics, implicit in normal linguistic usage. Second, as we will attempt to show, the historical evolution of the categorial structure of classical physics manifests an underlying consistency and progressive growth quite at variance with the scenarios proposed by paradigm replacements, degenerating research programs, problem-solving methodologies, and successive replacement of theories considered as separately interpreted units.
Surprisingly, the categorial structure of the language of classical physics has rarely been an object of systematic analysis. A schematic outline of the reasons for such neglect clarifies the issues to be addressed. A strong tradition in classical physics effectively treated the language of physics as a semi-transparent medium. Though opaque to the uninitiated, once mastered it serve as a medium through which one sees reality as it exists objectively. It seems reasonable, accordingly, to ignore the medium and focus on the reality revealed. Logical positivism rejected such naïve realism and made a sharp distinction between the language of functioning physics and the ideal languages proper to a rational reconstruction of physics. The latter involved formal languages for the rigorous reconstruction of theories and an observational language for the experimental reports that served to confirm or falsify theoretical predictions. Though the myth of a to-be-constructed observational language has long since vanished it left a perduring remnant, the designation of non-observational terms in experimental reports as 'theory laden'.
The reaction against positivism led to the semantic conception of theories. Its developers criticized syntactical reconstructions as being overly dependent on the particular linguistic formulation used in the axioms. Theories admit of diverse formulations, such as the Newtonian, Lagrangian, and Hamiltonian formulation of classical mechanics. Accordingly, it seemed advantageous to conceive of, or reconstruct, theories as abstract mathematical structures, such as the phase-space formulation of classical mechanics or the Hilbert-space formulation of quantum mechanics, interpreted through a family of models. One of these mathematical models is then taken as a model, in a more iconic sense, of some aspect of reality. In this conception, an interpretation of a theory is essentially a model-mediated relation between an abstract mathematical formulation and the things it is a theory of. Language effectively drops out of the final interpretation.