RESUMO Tradicionalmente, ser realista sobre algo significa crer na existência independente desse algo. Em termos ontológicos, isto é, acerca do que há, o realismo científico pode ser entendido como envolvendo a adoção de uma ontologia que seja cientificamente informada. Mas, segundo alguns filósofos, a atitude realista deve ir além da ontologia. A forma como essa exigência tem sido entendida envolve fornecer uma metafísica para as entidades postuladas pela ciência. Discutimos como duas abordagens em voga encaram o desafio de fornecer uma metafísica para a ciência: uma forma de naturalismo e a abordagem Viking/Toolbox. Por fim, apresentamos uma terceira via, que adota o melhor das duas abordagens: o método metapopperiano, que foca em descartarmos quais as alternativas erradas, ou melhor dizendo, os perfis metafísicos incompatíveis com certas teorias. Apresentamos o método metapopperiano, um método de metametafísica capaz de avaliar objetivamente quais os perfis metafísicos que são incompatíveis com certas teorias científicas. Para isso, usaremos como estudo de caso a mecânica quântica, mostrando resultados obtidos previamente. Com esse método, podemos ver como a ciência pode ser usada para evitar o erro em questões metafísicas. Essa seria, na nossa opinião, uma forma de desenvolver uma relação produtiva entre ciência e metafísica.

ABSTRACT Traditionally, being a realist about something means believing in the independent existence of that something. In this line of thought, a scientific realist is someone who believes in the objective existence of the entities postulated by our best scientific theories. In metaphysical terms, what does that mean? In ontological terms, i.e., in terms of what exists, scientific realism can be understood as involving the adoption of a scientifically informed ontology. But according to some philosophers, a realistic attitude must go beyond ontology. The way in which this requirement has been understood involves providing a metaphysics for the entities postulated by science, that is, answering questions about the nature of what ontology admits to exist. We discuss how two fashionable approaches face the challenge of providing a metaphysics for science: a form of naturalism and the Viking/Toolbox approach. Finally, we present a third way, which adopts the best of both approaches: the meta-Popperian method, which focuses on discarding the wrong alternatives, or better saying, the metaphysical profiles incompatible with certain theories. We present the meta-Popperian method, a metametaphysical method capable of objectively assessing which metaphysical profiles are incompatible with certain scientific theories. For this, we will use quantum mechanics as a case study, presenting some previously obtained results. As our focus is on methodological questions about the relationship between metaphysics and science; with this method, we can see how science can be used to avoid error in metaphysical issues. In our opinion, this would be a way to develop a productive relationship between science and metaphysics.

Abstract We articulate a distinction between ontology, understood as involving existence questions, and metaphysics, understood as either providing for metaphysical profiles of entities or else as dealing with fundamentality and/or grounding and dependence questions. The distinction, we argue, allows a better understanding of the roles of metaontology and metametaphysics when it comes to discussing the relations between ontology and science on the one hand, and metaphysics and science on the other. We argue that while ontology, as understood in this paper, may have reasonable perspectives for naturalization, given its relation to science, the same cannot be said for metaphysics, given that it is typically understood as an additional theoretical layer over science, not participating in the scientific investigation. That may result either in skepticism over metaphysics, or else on accepting that metaphysics is an autonomous branch of investigation, depending on one’s concern for metaphysics.