Standard rhetoric portrays science as self-correcting. Analysis of numerous historical cases helps clarify concretely when and how corrections in science occur, and when they do not. We may thereby articulate a general set of conditions necessary (although not necessarily sufficient) for when science may be considered self-correcting.
Philosophy of science has often privileged logic, method, and justification while neglecting a category that visibly structures scientific life: collective memory. This article argues that scientific rationality can be understood as a practice of living collective memory, grounded in Bergson's ontology of duration and memory. Bergson shows that the past does not simply disappear but endures virtually and can be actualized in response to present demands. Scientific knowledge can therefore be understood not merely as the accumulation of discrete results, but as a temporal process in which perception, habit, recollection, and invention cooperate. Halbwachs extends this problem into the collective register by showing that memory depends on social frameworks, institutions, languages, and shared temporal references. On this basis, the article develops living collective memory as a conceptual operator for philosophy of science, organized by three operations: conservation, reconstruction, and selective forgetting. The framework is tested through the Kuhn-Popper debate. Kuhn's account of paradigms clarifies how scientific communities conserve and reconstruct a usable past through exemplars, standards, and problem-solving routines. Popper's critical rationalism clarifies the complementary operation of selective forgetting, understood not as erasure but as the partial or complete withdrawal of a theory's authority to guide inquiry, while its traces remain available for historical, heuristic, pedagogical, or future reactivation. Ricoeur's reflections on forgetting and Derrida's analysis of the archive further show that scientific memory is selective, fragile, and institutionally mediated. The article does not reconcile Kuhn and Popper. It redescribes their opposition as a tension between two necessary functions of rationality: preserving continuity and withdrawing authority from what can no longer guide inquiry. Scientific progress, from this perspective, is neither simple accumulation nor pure rupture, but the regulated reorganization of a shared past. To treat science as living collective memory is to recognize rationality as fundamentally historical, collective, and precarious.
Otto Neurath's views on democracy, part of the largely forgotten political legacy of logical empiricism, emerged from his early theoretical reflections on human decision-making in terms of “auxiliary motives.” After further refinement during his engagement with Austromarxism and the Vienna Circle, it was only in his final years in Britain that Neurath articulated an explicitly democratic perspective, combining auxiliary motives with the idea of “orchestrating” politics within a framework of “scientific humanism.” This perspective remains strikingly relevant today, as it allows us to conceive of politics as both grounded in scientific and expert knowledge and responsive to the attitudes and preferences of ordinary citizens. This paper first provides a historical outline and then characterizes Neurath's views by introducing the notion of an unsystematic hierarchy of norms, which helps to demonstrate how auxiliary motives are necessary for arriving at concrete decisions within a framework of scientific humanism.
This paper develops a heuristic framework for understanding creative scientific judgment and demonstrates it through electrophysiology's development from the late 18th to early 20th century. I identify four patterns through which creative judgments manifest in scientific discovery: disruptive discovery, gap identification and boundary probing, technique-mediated discovery, and conceptual reframing. Drawing on Sánchez-Dorado's (2023) analysis, I argue that these creative judgments are characterised by 'pursuitworthiness'-their recognised potential to yield epistemic benefits if further explored. Historical episodes-Galvani's detection of animal electricity, du Bois-Reymond's instrumental innovations, Hermann's challenge to prevailing frameworks, and Bernstein's membrane theory-demonstrate how scientists navigated between established frameworks and their own extensions or revisions. This paper conceives scientific creativity as skilled judgment that illuminates overlooked aspects of epistemic tradition while articulating promising directions for future investigation.
When, why and how are scientific ideas worthy of pursuit has long been a subject of philosophical debates. This paper serves as an introduction to the Topical Collection "Pursuitworthiness in Science" by providing an overview of the literature on this topic, explicating the notion of pursuitworthiness judgments and by clarifying its applications to various contemporary discussions in philosophy of science and beyond.
What attitude should scientists have towards theories and theorizing? This paper criticizes one pervasive stance scientists take with regards to this question by assessing the significance of nonstandard genetic codes for Francis Crick's famous frozen accident theory (1968). When scientists engage with Crick's theory, they adopt the attitude that scientific theories are things to be confirmed or falsified. I argue this is the wrong attitude to adopt when assessing Crick's theory. I show how the logical complexity of Crick's theory has been widely misinterpreted. Not only has Crick's universality assumption been misunderstood, but the logical structure of frozen thesis lends itself to a plurality of (possibly incommensurable) operational specifications (Bich and Green 2018; Colaço 2022). A proper understanding of Crick's theory together with his own views on the purpose of theorizing suggests that something like provisional active scientific realism is the more appropriate attitude to take when assessing the frozen accident. Provisional active scientific realism is a logically weaker stance than Chang's (2014; 2022) view. On provisional active scientific realism, scientists should use Crick's theory as a vehicle for maximizing what they can learn about the world - at least at this early stage of investigation.
This paper focuses on an issue in the epistemology of interdisciplinary model transfer: how can model transfer between distinct domains of study make an epistemic contribution to science? I argue that there is an underappreciated but epistemically-significant difference between two modes of reasoning in model transfer, and which epistemic contributions model transfer can make depends on which mode is present. Existing discussions of interdisciplinary model transfers from one scientific discipline to another typically focus on analogy and analogical reasoning. By contrast, in what I'll call "type-token" reasoning, model transfer from one scientific system to another is motivated not by analogies or similarities but instead by the idea that different systems are tokens of a type. While analogies render phenomena more intelligible, type-token explanations aim at a certain kind of ontological unification.
In AI-driven science, there is an empirical asymmetry between AI researchers and experimenters, stemming from their different capabilities to control material resources and generate experimental data. This paper develops an "activity-centric approach to persuasion" by focusing on the case of AI-driven drug design to address the asymmetrical relationship between AI researchers and experimenters. First, AI researchers can emphasise several values to persuade experimenters to conduct epistemic activities for pursuing specific AI models. Second, "annotating" the outputs of AI models effectively facilitates the activities of experimenters. Finally, persuasion in AI-driven science proceeds through a reciprocal iteration of the activities between AI researchers and experimenters, which requires an analysis of the pursuit of an AI model at the organisational level. This activity-centric approach to persuasion would provide a useful framework for analysing interdisciplinary research projects, especially those involving asymmetrical relationships.
This article examines the use of computational tools in the history of science. For this reason, a historical case study is chosen: the interaction between Benjamin Franklin and Jean-Antoine Nollet over their competing theories of electricity in 18th-century France, focusing on the reception of Franklin's one-fluid theory in a Nollet-dominated Parisian Academy of Sciences. We built a database to study the reception and eventual acceptance of Franklin's theory in France. The historian of science John L. Heilbron portrays the Academy as divided and paralyzed, while Roderick W. Home argues that Nollet's dominance led the Academy to conduct the debate in Nollet's shadow. Using a database of publications on electricity related to the Parisian Academy between 1745 and 1785, we examined Franklin's theory in France quantitatively, revealing a pattern of influence and intellectual dominance that changed abruptly after Nollet's death. We also produced data comparing Nollet and another important figure in 18th-century French electricity, Jean-Baptiste Le Roy, and reconstructed this debate using statistics. Our findings demonstrate that databases and scientometric methods can be fruitfully applied in the history and philosophy of science. In particular, we can confirm part of the dynamics of this historical event, especially related to Nollet's influence within the Parisian Academy, an extra-scientific factor. For further exploration of this case study and to test these tools more thoroughly, a larger database of electricity-related publications in France is envisioned as the next step.
This paper proposes a reinterpretation of predictive coding (PC) in terms of heterarchical networks of control mechanisms. Standard accounts of PC typically assume a rigid hierarchical organization, in which information flows in fixed directions across levels of abstraction. Such a framework has, however, faced growing theoretical criticism and appears increasingly difficult to reconcile with recent research on predictive routing, which suggests that predictive systems exhibit greater flexibility than hierarchical models allow. In response to these challenges, we propose integrating work on PC with both complex network theory and William Bechtel's conception of control mechanisms. Our central hypothesis is that predictive systems can be understood as emergent patterns arising from the interaction between hierarchical production mechanisms and heterarchical networks of control mechanisms. On this view, local computational hierarchies are dynamically modulated by flexible contextual constraints, thereby preserving the advantages of hierarchical models while avoiding their rigidity. The paper is theoretical in nature and highlights the need for empirical validation of the proposed framework, as well as for a systematic integration of network methodologies with research on predictive processing.
In this paper, we investigate the relationship between Popper's propensity interpretation of probability and Bohr's principle of complementarity. While Popper is often regarded as a philosophical antagonist of the Copenhagen interpretation, and Bohr never explicitly articulated his views on the nature of probability, we propose that the propensity interpretation can provide a coherent ontological foundation for Bohr's complementarity principle. This paper is primarily a rational reconstruction rather than a historical claim about Bohr's explicit commitments. We argue that both propensities and complementarity point toward a similar ontological picture-a move away from an object-based ontology of substances possessing intrinsic properties toward an ontology grounded in interactions, dispositions, and processes. According to this view, quantum phenomena are not predetermined but emerge from propensities inherent in the complete experimental arrangement. We also briefly consider how this framework relates to modern approaches of the decoherence-based program of emergent classicality.
Mechanistic theories of explanation are widely held in the philosophy of science, especially in philosophy of biology, neuroscience and cognitive science. While such theories remain dominant in the field, there have been an increasing number of challenges raised against them over the past decade. These challenges claim that mechanistic explanations can lead to incoherence, triviality, or deviate too far from how scientists in the life sciences genuinely employ the term "mechanism". In this paper, I argue that these disputes are fueled, in part, by the running together of distinct questions and concerns regarding mechanisms, representations of mechanisms, and mechanistic explanation. More care and attention to how these are distinct from one another, but also the various ways they might relate, can help to push these disputes in more positive directions. I then highlight the sort of philosophical problems and concerns that we are better served focusing our attention on instead.
I pose the epistemological question of what makes the transfer of statistical approaches across disciplines, specifically between physics and biology, legitimate and fruitful despite intrinsic differences in their objects of study - a problem that resurfaces in contemporary interdisciplinary research relying on machine learning for statistical model building. I address it through the historical reconstruction of pivotal steps in the development of statistical thinking in the 19th century, where the appeal to the mathematical formalism of the Gaussian distribution acted as the visible trace of the diffusion of statistical approaches from astronomy to biological and social sciences. My analysis positions the wide-reaching, nowadays accepted applicability of statistics as something historically acquired through gradual conceptual and technical elaboration. It expounds the forms of re-sanctioning that accompanied and enabled the cross-domain transfer of the statistical approach, articulating them in terms of re-interpretation of the mathematical descriptions involved, re-formulation of the underlying assumptions, and re-conceptualization of their theoretical status and foundations from theory-related abstractions to approximations. The latter culminated in a shift of attitude that led to perceiving, as is standard nowadays, statistical mathematical descriptions as convenient tools for quantitative analysis, further legitimating and accelerating their interdisciplinary transfer. This work aims to familiarize historians and philosophers of science, as well as physicists, mathematicians and biologists with an interest in the history of their discipline, with these key episodes, and to dissect the epistemological assumptions and implications, as well as the interpretive frameworks, at stake in the application of a statistical approach across disciplines.
Quantum gravity suggests that spacetime may not be fundamental, and it has been argued that we can understand a theory without a fundamental spacetime if we are able to claim that spacetime `emerges' from some non-spatiotemporal entities. In this sense, strategies like functionalism have been deployed to claim that this emergence is possible and plausible, both in principle and in practice for current approaches to quantum gravity. In this article I argue that this analysis is incomplete, as it tends to overlook the way the dynamics of these theories is `quantum' in a way that differs from standard quantum theory. The challenge for the emergentist (and for the quantum gravity theorist) is to give an interpretation not only to the kinematical and classical aspects of these theories, but to the dynamical and quantum ones, and to show how the spacetime roles can be fulfilled, if possible at all. Therefore, I argue that some current approaches to quantum gravity seem to fail to provide meaningful theories, that spacetime functionalism is of no help, and that the position of the spacetime emergentists is weakened, as they lack any example of a successful reduction of spacetime to some truly quantum non-spatiotemporal stuff.
The main aim of this article is to significantly expand the frame model in order to analyze scientific theories by means of what I call theory frame modules (TFMs). I will illustrate the notion of TFMs through three scientific theories (Aim 1), each employing different scales of measurement that will be analyzed by corresponding TFMs (Aim 2). To this end, I will propose a frame-based definition of theoretical concepts (Aim 3). In the next step, I will focus on intertheoretical relations between TFMs and provide frame-based definitions of the relations of specialization and theoretization (Aim 4). Frames offer a means of reconstructing recursive structures. To provide a more differentiated perspective on recursion within frames, I will distinguish between recursion in the narrow sense and recursion in the broad sense (Aim 5). Finally, I will propose a frame-based distinction between an operationalist and a unificationist approach to scientific concepts, the latter of which allows for the introduction of theoretical concepts that designate a common cause of different empirical phenomena (Aim 6).
It is a well-known predicament of the social sciences that predictions can sometimes intervene on the very processes the predictions concern. The purpose of this paper is to address the question of how modelers or scientists should act in the presence of such performative effects with a specific focus on the example of climate economics. Specifically, we will argue that two strategies for managing performative effects recently defended in the literature (i.e., mitigating and appraisal) will not be adequate for all cases because we often lack the knowledge to execute them properly. We will focus on the use of so-called integrated assessment models (IAMs) in climate policy, where the policy process spans large time-horizons and heavily depends on a diverse set of social actors. For cases like this, we argue that we should do neither mitigating nor appraisal and propose that it can be adequate to act as if one does not anticipate any performative effects.
