Bruno Malavolta e Silva

Resumo: O argumento da subdeterminação constitui um dos principais argumentos contra o realismo científico. Analiso diversas versões do argumento, e defendo que ele se torna mais plausível quando entendido como um argumento indireto contra o realismo. Tal proposta requer distinguir entre três maneiras principais de formular o argumento da subdeterminação. Na formulação tradicional, o argumento baseia-se na formulação de teorias rivais que sejam empiricamente adequadas à evidência disponível. Na formulação kuhniana, o argumento baseia-se na inexistência de um algoritmo neutro de normas epistêmicas que seja sufi ciente para determinar a escolha de teorias. Tais formulações do problema da subdeterminação são amplamente rejeitadas enquanto ameaças globais ao realismo científico. Proponho que, alternativamente, a relevância de fatores não epistêmicos poderá ser melhor apreciada pelo realismo se focarmos no impacto indireto que tais fatores possuem na escolha de teorias, em vez de focarmos no fato de as normas epistêmicas não serem neutras. Com isso, o foco do argumento da subdeterminação deverá ser o de problematizar como fatores do contexto de descoberta influenciam indiretamente a objetividade do processo de justificação das teorias científicas, na medida em que modificam a evidência e hipóteses disponíveis aos cientistas em seu ambiente epistêmico. Palavras-Chave: Ciência e valores. Objetividade científica. Realismo científico. Subdeterminação da teoria pela evidência.

O argumento do milagre afirma que o realismo científico é a melhor explicação para o sucesso da ciência: teorias científicas são bem-sucedidas porque são verdadeiras, e cientistas são bem-sucedidos em encontrar teorias verdadeiras porque se baseiam em normas metodológicas confiáveis. A tese da incomensurabilidade metodológica afirma que teorias científicas não são escolhidas através de um algoritmo neutro de normas epistêmicas. Isso revela uma lacuna na explicação realista: normas epistêmicas confiáveis não são suficientes para conduzir a escolhas de teorias verdadeiras, pois tais escolhas também são determinadas por outros fatores além de normas epistêmicas. A introdução de fatores adicionais como subdeterminando a escolha de teorias engendrou argumentos relativistas e antirrealistas contra o realismo. A explicação realista pode ser reabilitada se postular que os cientistas possuem autoridade epistêmica para aprimorar as normas metodológicas da ciência, sendo aptos a tomar decisões autônomas em vez de regidas por regras.

Stance voluntarism highlights the role of the will in epistemic agency, claiming that agents can control the epistemic stances they assume in forming beliefs. It claims that radical belief changes are not compelled by the evidence; they are rationally permitted choices about which epistemic stances to adopt. However, terms like "will", "choice", and "stance" play a crucial role while being left as vague notions. This paper investigates what kind of control rational agents can have over epistemic stances. I argue that whether epistemic stances are voluntary depends on what kind of stance is being assessed. Sometimes epistemic stances are taken to be evaluative attitudes about how to produce knowledge. This kind of stance is not directly controllable, since it is essentially connected to beliefs, and believing is not voluntary. But sometimes epistemic stances are taken to be styles of reasoning and modes of engagement, expressing ways of approaching the world in order to produce knowledge, which can be voluntary. Overall, this supports a formulation of stance voluntarism as a dual-systems theory of epistemic agency, where epistemic rationality is compounded by a dynamic interplay between involuntary processes of belief formation and voluntary processes of cognitive guidance.

The problem of unconceived alternatives (or the New Induction) states that, since scientists have recurrently failed to conceive relevant theoretical alternatives for some domains of science, current scientists are probably also failing to do so. Therefore, there may be theories which still exceed the grasp of scientists' imagination, and one should not endorse a realist stance towards current science. In this paper, I raise a conceptual worry for the formulation of this problem: what does it mean to say that scientists failed to conceive a relevant theory? What aggravates the problem is that no simple notion of relevance makes the New Induction as strong as it initially seems. I consider the three more obvious interpretations of relevance: relevance as objective probability; relevance as epistemic probability assessed by current scientists; and relevance as epistemic probability assessed by past scientists. I argue that assuming any of these three notions implies difficulties for the New Induction, hence their proponents shouldn't take the notion of relevance for granted. A more precise definition of relevance is essential to understand what are the difficulties surrounding the problem of unconceived alternatives as an epistemic worry. Until now, such notion is missing.

Antirealist explanations for the success of science have been widely discussed up to today and have received several formulations. This makes it rather complex to assess them all. The objective of this paper is to help understand and assess the proposal of an anti-realist explanation for science's success. I show the core assumptions contained in the several anti-realist explanations, how they relate to each other, and which background assumptions are required in order to warrant each position. I argue that, since the many anti-realist explanations are only plausible when maintained conjointly, there is essentially only one anti-realist account to science's success: scientists find successful (and even fertile) theories because they use methods of theory-selection and theory-construction that preserve only successful theories; the fact that these theories prove themselves successful will not be mysterious if it is conjointly assumed that false theories are often empirically successful. This explanation relies on a semantic and a methodological view concerning the probabilistic relation between success and truth, and also on an epistemic stance regarding the limits of explanatory reasoning. The crucial divergence between realist and antirealist accounts of science's success lies in how probable they assess the possibility of a theory to be false and empirically successful. Since the stalemate between realist and antirealist explanations results from a prior disagreement about the probabilistic connection between success and truth (or the underdetermination thesis), the challenge raised by the antirealist explanation to realism becomes equivalent to the traditional charge that the no-miracles argument is circular.

Arthur Fine presented the Natural Ontological Attitude (NOA) as a third alternative between scientific realism and anti-realism by identifying a core position contained in both and rejecting any philosophical addition to this core. At first, Fine's proposal was understood as offering a doxastic middle ground between believing in the truth of a theory and believing in its empirical adequacy. In this reading, NOA was widely disregarded after Alan Musgrave's criticisms of it, which characterized Fine's proposal as a form of realism. After that, NOA was reinterpreted as a local variety of realism focused in changing the attitude used to settle the scientific realism debate, by rejecting global philosophies with an approach external to science, and by considering only the scientific evidence with a contextualist mood. Although this reading clarifies how to understand NOA, there is still no consensus about what is Fine's argument to support it. I organize the four main interpretations of Fine's defense and point their main flaws. Finally, I develop some clarifications about NOA in order to solve the flaws of the preceding interpretations, defending that NOA is based upon a prevalence of the epistemic values actually used in scientific practice. 1. Introdução "O Realismo está morto", declara Arthur Fine sobre o realismo científico. Sua morte é anunciada por físicos que desenvolvem sua ciência sem aderir ao realismo, e por neo-positivistas que aceitam todos os resultados científicos sem sequer precisar considerar questões de realismo como mais do que pseudoquestões. Tentativas de defender o re-alismo perduram, mas representam o primeiro estágio do luto, o estágio de negação (Cf. Fine 1986a, p.83). Fine vê o debate sobre realismo científico como estagnado e infértil. Para superar a estagnação desse debate, Fine oferece a Atitude Ontológica Natural como uma al-ternativa teórica para uma filosofia da ciência progressiva e que não se comprometa com realismo ou anti-realismo (Cf. Fine 1986a, p.84). A NOA (Natural Ontological Attitude) 1 é constituída por dois elementos assumidos conjuntamente: a posição central ; e o anti-essencialismo filosófico. Deixe-me explicá-los.

Scientific Realism has been defended almost exclusively by the Non-Miracle Argument, which states that either it‘s assumed that scientific theories are true or the success of the scientific enterprise becomes a mysterious and unexplainable fact. At first, the main criticisms of the Non-Miracle Argument are gathered in five objections: (i) it‘s presented an alternative explanation to the success of science that competes with the realist explanation and does not assume the truth of the scientific theories; (ii) it‘s defended that the realist hypothesis is not a satisfactory explanation to the success of science, charging it of being an empty explanation, or (iii) charging its capacities to reach the scientific methodological rigorousness that itself imposes to the warrant of theories; (iv) it‘s affirmed that the argument is a petition principii, on account of presupposing an abductive rule of inference accepted only by the realist perspective; (v) it‘s proposed that the argument constitutes a statistical fallacy on account of neglecting a relevant base rate; When replying to the presented objections, some authors come up with the distinction between a semantic version of the argument, based on the empiric success reached by scientific theories considered as semantic entities, and a methodological version of the argument, based on the pragmatic success of scientists at choosing fruitful theories and discovering succeeded theories. Furthermore, other authors offer a second distinction between a general version of the argument, statistically based on the generalized success of science, and a retail version of the argument, based on its specific application to an individual theory. After comparing the advantages and drawbacks of each interpretation, the relevance of each distinction is reanalyzed and a new version of the argument by making a synthesis of its different interpretations is defended in a way to answer the five initial objections.

Resumo: O realismo entidades de Ian Hacking é defendido como um meio-termo no debate entre realismo e anti-realismo científico. A maior parte da literatura interpreta o realismo de entidades equiparando-o ao realismo fenomenológico de Nancy Cartwright e focando-se no argumento experimental ou argumento pela intervenção oferecido por Hacking. Neste artigo, defende-se a relevância de uma segunda dimensão do realismo de entidades, segundo a qual estamos justificados a afirmar a existência de entidades observadas instrumentalmente, independentemente das leis fenomenológicas associadas a elas. Separando-se o papel da manipulação causal e repensando o papel da observação instrumental, obtém-se uma compreensão mais clara dos argumentos de Hacking em favor da microscopia, tal como uma posição mais compatível com sua filosofia da experimentação. Palavras Chave: Realismo Científico; Realismo de Entidades; Hacking; Representar e Intervir.

The explanatory defense of scientific realism is motivated by science’s success. By contrast, local realists claim that scientific realism must be justified contextually, by considering individually the scientific evidence relevant to assess each theory. I propose that both approaches can be harmonized by appealing to a particularist epistemology. The characterization of local realism as a form of particularism allows us to re-evaluate its relation to the explanatory defense. Local realists reject the explanatory defense as incompatible with the localist stance, but according to particularism, our attribution of knowledge to particular cases is made prima facie, in a way open to the refinement of our epistemic norms. The function of epistemic criteria is not to demonstrate that we have knowledge, but rather to reveal its epistemic nature. One can formulate epistemological criteria by inspecting particular instances of knowledge. And one can evaluate the adequacy of his criteria by adjusting them to his cognitive life in a process of reflective equilibrium. And this is precisely what the explanatory defense of realism does. In this thesis, I develop a particularist defense of scientific realism, showing how a realist stance can be reasonably sustained in the face of the main anti-realist arguments: the pessimistic meta-induction; the problem of unconceived alternatives; and the argument from underdetermination of theory by data.