Rudolf Carnap’s Empiricism, Semantics, and Ontology addresses an objection raised by empiricist and nominalist critics against formal semantics: that semantics, because it speaks of properties, classes, numbers, propositions, and other abstract entities, covertly commits one to a Platonic ontology and is therefore incompatible with empiricism. Carnap’s central aim is to show that using a language that refers to abstract entities does not by itself commit one to a metaphysical thesis about their independent reality. The work argues that the apparent force of the objection comes from a failure to distinguish different kinds of existential question.
Carnap introduces the notion of a linguistic framework: a system of rules that determines what expressions may be used and what counts as a correct answer to questions posed within it. Once a framework has been adopted, questions about the existence of entities within it—internal questions—can be answered by logical or empirical procedures, depending on the kind of framework. In the everyday “thing language,” the question whether there is a sheet of paper on a desk is empirical. In the framework of numbers, the question whether there is a prime number greater than a given number is settled by logical proof. Such internal questions are genuine cognitive questions, and their answers are governed by the rules of the framework.
External questions, by contrast, concern the existence or reality of the framework as a whole: for example, “Are there really physical objects?” or “Do numbers really exist?” Carnap argues that if such questions are intended as theoretical questions about a reality lying behind the framework, they are pseudo-questions. There is no neutral, super-scientific standpoint from which one could compare the framework as a whole with the world and pronounce on its ultimate reality. The only meaningful question in the vicinity is a practical one: whether to adopt and use the framework. That decision is to be judged by the framework’s efficiency, fruitfulness, simplicity, and usefulness for science and communication, not by evidence for a metaphysical thesis.
Carnap applies this distinction to the traditional dispute between realism and idealism about the external world. The realist’s “yes” and the idealist’s “no” to the question whether the world of things really exists are not theoretical answers supported by evidence; both parties in practice use the thing language. The theoretical external question is not answerable, while the practical acceptance of the thing language is justified by its utility. Similarly, in the dispute between nominalism and Platonism about abstract entities, the nominalist’s rejection of abstract entities is best understood as a recommendation about what language to use, not as a factual discovery. Semantics may legitimately use abstract entities as part of a linguistic framework; the empiricist need not fear that this amounts to embracing a Platonic ontology. The acceptance or rejection of abstract linguistic forms is a practical decision about the instruments of scientific and logical work.
Carnap also clarifies that “real” is framework-relative. Within a framework, calling something “real” contrasts it with what is fictitious, imaginary, or merely apparent. The term cannot be applied in the same sense to the framework itself. The essay therefore rejects the idea of ontology as a super-scientific discipline that prescribes to science which entities are admissible. Its conclusion is that empiricism can consistently use semantics and abstract entities, provided that the metaphysical interpretation of those entities is abandoned. The argument is not a verificationist refutation of Cartesian skepticism; it treats external theoretical questions about the reality of a framework as pseudo-questions and redirects attention to the practical choice of linguistic frameworks.
A further theme is the distinction between accepting a framework and asserting something within it. Accepting a framework is a matter of adopting rules for forming and transforming sentences, and thereby fixing what counts as a meaningful question and an admissible answer. This acceptance is not itself a claim made inside the framework; it is a practical decision about a linguistic instrument. Once the framework is in use, however, internal existence claims are objective relative to its rules—so “there are numbers” or “there are physical objects” can be trivially or empirically true within the relevant framework, while the metaphysical external question has no cognitive content.
The essay also defends semantic analysis against nominalist objections by treating semantic notions—such as expression, designation, and truth—as requiring a framework in which abstract entities may be used. Carnap’s point is not to assert a Platonic doctrine but to allow such entities as part of the machinery of logical and semantic inquiry. The empiricist can adopt this machinery without ontological guilt, because the question is whether the framework is useful for logic, mathematics, and science, not whether it mirrors a mind-independent realm of abstracta.
The resulting outlook is pluralist and pragmatic: there is no unique true ontology, and different linguistic frameworks may be appropriate for different purposes. Philosophy’s task is not to legislate what exists but to clarify the rules of frameworks and evaluate them by their fruitfulness. The essay’s significance lies in this reframing of ontological disputes as disputes about language and policy, and in showing that empiricism can be tolerant of abstracta while remaining anti-metaphysical.
