to the evolution of contemporary pragmatism philosophers had identified realism as such with one or another particular ontology, which they erroneously viewed as the only ontology on the assumption that there can be only one ontology. Such is the error made by some physicists who believe that they are defending realism, when they defend the “hidden variable” interpretation of quantum theory. Such too is the error in Popper’s proposal for his propensity interpretation of quantum theory. As it happens, both Bohm and Heisenberg rejected the ontological thesis that the kind of existence familiar to us can be extrapolated into the atomic order of magnitude. And contrary to Einstein’s EPR thesis of a single uniform ontology for physics, Aspect, Dalibard, and Roger’s findings from their 1982 nonlocality experiments demonstrated empirically the Copenhagen interpretation’s semantics and ontology.
Advancing science has produced revolutionary changes. And as the advancement of science has produced new theories with new semantics exhibiting new ontologies, some prepragmatist scientists and philosophers found themselves attacking a new theory and defending an old theory, because they had identified realism with the ontology associated with the older falsified theory. As Feyerabend notes in his Against Method, scientists have criticized a new theory using the semantics and ontology of an earlier theory. Such a perversion of scientific criticism is still common in the social sciences where romantic ontologies are invoked as criteria for criticism.
With ontological relativity realism is no longer uniquely associated with any one particular ontology. The ontological-relativity thesis does not deny metaphysical realism, but depends on it. It distinguishes the mind-independent plenitude from the ontologies revealed by the descriptive semantics of more or less empirically adequate beliefs. Ontological relativity enables admitting change of ontology without resorting to instrumentalism, idealism, phenomenalism, solipsism, any of the several varieties of antirealism, or any other such denial of metaphysical realism.
Thus ontological relativity solves the modern problem of reconciling conceptual revision in science with metaphysical realism. Ontological relativity enables acknowledging the creative variability of knowledge operative in the relativized semantics and consequently mind-dependent ontology that are defined in constructed theories, while at the same time acknowledging the regulative discipline of mind-independent reality operative in the empirical constraint in tests with their possibly falsifying outcomes.
Or as manifested linguistically, using the relativized semantics and ontology partially defined by the theory “Every X is Y” it is possible to articulate the falsifying observational test outcome “This X is not Y” with the semantics defined by the test-design language, even though after issuing the falsifying observational statement the consequent revision of belief in the theory “Every X is Y” occasions a partial change in the rejected theory’s descriptive semantics and its ontological claim.
In contemporary pragmatist philosophy of science metaphysical realism is logically prior to and presumed by all ontologies as the primal prejudice, while the choice of an ontology is based upon the empirically demonstrated adequacy of the theory describing the ontology. Indulging in futile disputations about metaphysical realism will not enhance achievement of the aims of either science or philosophy of science, nor will dismissing such disputations encumber achieving those aims. Thus ontological relativity leaves ontology to the scientist rather than the metaphysician. And the superior empirical adequacy of a new law yields the increased truth of a new law and the increased realism in the ontology that the new law reveals.
3.38 Ontological Relativity Illustrated
To illustrate ontological relativity consider the semantical decision about red ravens mentioned in the above discussion about componential artifactual semantics (Section 3.23). The decision is ontological as well as semantical. For the bird watcher who found a red but otherwise raven-looking bird and decides to reject the belief “Every raven is black”, the phrase “red raven” becomes a description for a type of existing birds. Once that semantical decision is made, red ravens suddenly populate many trees in the world, however long ago Darwinian Mother Nature had evolved the observed avian creatures. But if his decision is to persist in believing “Every raven is black”, then there are no red ravens in existence, because whatever kind of creature the bird watcher observed and that Mother Nature had long ago evolved, the red bird is not a raven. The availability of the choice illustrates the artifactuality of the relativized semantics of language and the consequently relativized ontology that the relativized semantics reveals about mind-independent reality.
Relativized semantics makes ontology no less relative whether the affirmed entity is an elephant, an electron, or an elf. Beliefs that enable us routinely to make successful predictions are deemed more empirically adequate and thus more realistic and truer than those less successfully predictive. And we recognize the reality of the entities, attributes or any other characteristics that enable those routinely successful predicting beliefs. Thus if positing evil elves conspiring mischievously enabled predicting the collapse of market-price bubbles more accurately and reliably than the postulate of euphoric humans speculating greedily, then we would decide that the ontology of evil elves is as adequately realistic as it was found to be adequately empirical, and we would busy ourselves investigating elves, as we would do with elephants and electrons for successful predictions about elephants and electrons. On the other hand were our price predictions to fail, those failures would inform us that our belief in elves is as empirically inadequate as the similarly discredited belief in the gnomes of Zurich, and we would decide that the ontology of elves is as inadequately realistic, as it was found to be inadequately empirical.
Consider another illustration. Today we reject an ontology of illnesses due to possessing demons as inadequately realistic, because we do not find ontological claims about possessing demons to be empirically adequate for effective medical practice. But it could have been like the semantics of “atom”. The semantics and ontology of “atom” have changed greatly since the days of the ancient philosophers Democritus and Leucippus. The semantics of “atom” has since been revised repeatedly under the regulation of empirical research in physics, as when J.J. Thompson discovered that the atom is not indivisible, and thus today we still accept a semantics and ontology of atoms. Similarly the semantics of “demon” might too have been revised to become as beneficial as the modern meaning of “bacterium”, had empirical testing regulated an evolving semantics and ontology of “demon”.
Both ancient and modern physicians may observe and describe some of the same symptoms for a certain infectious disease in a sick patient and both demons and bacteria are viewed as living agents, thus giving some continuity to the semantics and ontology of “demon” through the ages. But physicians’ medical understanding, diagnoses and remedies are quite different. If the semantics and ontology of “demon” had been revised under the regulation of increasing empirical adequacy, then today scientists might materialize (i.e., visualize) demons with microscopes, physicians might write incantations (i.e., prescriptions), and pharmacists might dispense antidemonics (i.e., antibiotics) to exorcise (i.e., to cure) possessed (i.e., infected) sick persons. But then terms such as “materialize”, “incantation”, “antidemonics”, “exorcise” and “possessed” would also have acquired new semantics in the more empirically adequate modern contexts than the ancient medical beliefs. The descriptive semantics and ontology of “demon” would have been revised to exclude what we now find empirically to be inadequately realistic, such as a demon’s immateriality.
This thesis can be found in Quine’s “Two Dogmas of Empiricism” in his Logical Point of View even before he came to call it “ontological relativity”. There he says that physical objects are conceptually imported into the linguistic system as convenient intermediaries, as irreducible posits comparable epistemologically to the gods of Homer. But physical objects are epistemologically superior to other posits including the gods of Homer, because the former have proved to be more efficacious as a device for working a manageable structure into the flux of experience. As a realist, he might have added explicitly that experience is experience of something, and that physical objects are more efficacious than whimsical gods for making correct predictions.
Or consider the tooth-fairy ontology. In some cultures young children losing their first set of teeth are told that if they place