Jinyeong Gim

After Wesley Salmon’s causal-mechanical stance on explanation in the 1980s, the ontic-epistemic debate of scientific explanations appeared to be resolved in the philosophy of science. However, since the twenty-first century, this debate has been rekindled among philosophers who focus on mechanistic explanations. Nevertheless, its issues have evolved, necessitating scrutiny of the new trends in this debate and a comparison with the original controversy between Carl Hempel and Salmon. The primary objective of this paper is to elucidate three categorical dimensions in the ontic-epistemic debates, spanning from the original to the recent controversies. Subsequently, it will explore why the conception of explanation is linked to representations, what conditions are necessary for linguistic expressions to be explanatory, and what roles norms play in explanation. Consequently, contrary to the common stereotype, it will be argued that mechanistic explanations are more likely to be epistemic rather than ontic.

The New Mechanism has traditionally focused on identifying constitutively relevant components in mechanistic explanation but has largely overlooked how temporal organization encompassing order, duration, and rate should be formally represented. This paper develops a structuralist approach to mechanistic explanation by integrating insights from scientific representation and category theory. Drawing on Hughes’ DDI (Denotation, Demonstration, Interpretation) account, it establishes a representational framework for modeling temporal constraints as explanatory targets. Category–theoretic tools including index categories, functorial mappings, and Petri nets are employed to formally encode dynamical and sequential structures. Using protein synthesis as a case study, the paper demonstrates that mechanistic explanation must account not only for token–level causal dependencies but also for type–level structural constraints. By bridging the theory of scientific representation with mechanistic explanation, it clarifies how temporal organization underlies and enables explanatory models in biology.

This paper offers a Bayesian critique of Joshua Greene’s dual-process model of moral judgment, which implicitly posits a structural division between emotional and cognitive systems. Drawing on recent developments in computational neuroscience, particularly predictive coding, active inference, and constructionist theories of emotion, I argue that both emotional and cognitive functions emerge from a unified inferential system that minimizes prediction error across bodily and environmental domains. This view suggests that the functional duality observed in emotional and cognitive moral judgments does not necessarily imply a structural dualism in the brain’s computational system. Reframing moral judgment within a monolithic Bayesian framework dissolves traditional dichotomies such as reason versus emotion, and reconceives moral judgment as an embodied, context-sensitive process of predictive adaptation. This shift has significant implications for neuroethics and moral psychology, inviting a move beyond static dual process models toward a dynamic, integrative account of moral judgment.

This paper explores moral judgments as conscious phenomena, highlighting the necessity of an interdisciplinary approach to achieve a comprehensive scientific understanding. Integrating insights from neuroscience, moral psychology, and philosophy of science, it adopts a mechanistic framework to analyze moral judgments through three dimensions: informational diversity, functional dynamics, and intentionality. While the mechanistic approach provides a robust foundation for decomposing moral judgments into their constitutive elements, it faces limitations in establishing causal relationships across neural levels and integrating complex cognitive processes. To address these challenges, it evaluates neuroscientific theories of consciousness, identifying Global Workspace Theory (GWT) as the most suitable framework for unifying distributed neural processes into coherent conscious experiences. By synthesizing mechanistic philosophy with GWT, this paper offers a novel explanatory model for understanding moral judgments as multidimensional, integrative, and conscious phenomena.

Craver's diagram, comprising symbols such as X (entity), S (mechanism), Φ (activity), and Ψ (phenomenon), is widely used to represent biological mechanisms in the New Mechanism. However, this paper demonstrates that Craver’s framework lacks the formal capacity to adequately capture the organizational structures and functional dynamics essential for mechanistic explanations, particularly the temporal interplay among entities and activities or the relational nature of enzymatic state transitions. To address these limitations, this paper proposes a supplementary framework based on category theory, enabling the abstraction of mechanisms as temporally organized structures of state transitions. By integrating key features such as order, frequency, and duration, categorical diagrams provide a cohesive representation of type mechanisms, supporting the deficiencies of token-centric approaches. Using protein synthesis as a case study, the paper illustrates how categorical abstractions enhance the formal representation of biological mechanisms while emphasizing the philosophical importance of organizational dimensions.

This paper examines two prominent frameworks in moral psychology that aim to explain the foundations of moral judgments: the Universal Moral Grammar (UMG) and the Emotional Moral Sentimentalist (EMS) frameworks. UMG theorists propose that moral judgments are guided by innate, unconscious principles analogous to linguistic grammar, whereas EMS advocates emphasize the role of automatic emotional responses, such as anger or guilt, in shaping moral decisions. By reviewing recent neuroscientific evidence, this paper evaluates the strengths and limitations of each approach and argues that neither framework is fully substantiated by current empirical findings. Instead, it advocates for an integrative perspective that treats UMG principles and emotional processes as complementary, hypothetical models for understanding moral cognition. This approach offers a pathway for more comprehensive investigations into the neural mechanisms underlying moral judgments, addressing the limitations of the ongoing UMG-EMS debate.

After Wesley Salmon’s causal-mechanical stance on explanation in the 1980s, the ontic-epistemic debate of scientific explanations appeared to be resolved in the philosophy of science. However, since the twenty-first century, this debate has been rekindled among philosophers who focus on mechanistic explanations. Nevertheless, its issues have evolved, necessitating scrutiny of the new trends in this debate and a comparison with the original controversy between Carl Hempel and Salmon. The primary objective of this paper is to elucidate three categorical dimensions in the ontic-epistemic debates, spanning from the original to the recent controversies. Subsequently, it will explore why the conception of explanation is linked to representations, what conditions are necessary for linguistic expressions to be explanatory, and what roles norms play in explanation. Consequently, contrary to the common stereotype, it will be argued that mechanistic explanations are more likely to be epistemic rather than ontic.

The type-level abstraction is a formal way to represent molecular structures in biological practice. Graphical representations of molecular structures of biological objects are also used to identify functional processes of things. This paper will reveal that category theory is a formal mathematical language not only to visualize molecular structures of biological objects as type-level abstraction formally but also to understand how to infer biological functions from the molecular structures of biological objects. Category theory is a toolkit to understand biological knowledge at the type-level formally, not individual token-level, as well as typical heuristic strategies in molecular biology.

Leonelli (2016) suggested a relational view of data against a representational view by emphasizing data-centric biology rather than the theory-centric tradition in the philosophy of science. This is because the first view allows for data journeys across laboratories using public database resources, whereas the second does not. This paper examines Leonelli’s strategies to defend the relational view of data. Contrary to Leonelli’s intention, it indicates that her strategies led to unnecessary misunderstandings of the relationships among data, representation, and evidential values. It will be argued that evidential values of data are inevitably based on a representational feature and that it is better to reconcile both views of data as complementary rather than contrasting.

South Korean high school students are being taught Einstein’s Special Theory of Relativity. In this article, I examine the portrayal of this theory in South Korean high school physics textbooks and discuss an alternative method used to solve the analyzed problems. This examination of how these South Korean textbooks present this theory has revealed two main flaws: First, the textbooks’ contents present historically fallacious backgrounds regarding the origin of this theory because of a blind dependence on popular undergraduate textbooks, which ignore the revolutionary aspects of the theory in physics. And second, the current ingredients of teaching this theory are so simply enumerated and conceptually confused that students are not provided with good opportunities to develop critical capacities for evaluating scientific theories. Reviewing textbooks used in South Korea, I will, first, claim that the history of science contributes to understand not merely the origins but also two principles of this theory. Second, in addition to this claim, I argue that we should distinguish not only hypotheses from principles but also phenomena from theoretical consequences and evidence. Finally, I suggest an alternative way in which theory testing occurs in the process of evaluation among competitive theories on the basis of data, not in the simple relation between a hypothesis and evidence.