Conference Talks
Khan, A. (2026). “Robustness Analysis and the Energy Balance model of Obesity,” Canadian Society for the History and Philosophy of Science (CSHPS). Halifax, NS, Canada.
Abstract: This talk reassesses the evidential role of robustness analysis (RA) in model-based science through a case study of obesity modelling, focusing on Hall et al.’s (2020) Energy Balance Model (EBM). I argue that the standard RA taxonomies—within-framework, across-framework, and explanatory RA (ERA)—presuppose a strict model–target boundary and therefore cannot, on their own, underwrite world-directed confirmation in this domain. Within-framework RA remains diagnostic but intra-model; across-framework RA expands coverage but inherits incompatible idealizations; and ERA’s eliminative logic falters unless one assumes a contentious conjunctive adequacy bridge from model detections to worldly facts. Drawing on Hall et al.’s model, and the literature that they survey, I show that the explanandum “obesity” is iteratively re-specified alongside its model. I thus propose “iterative model–target co-construction” as the right way to understand this case. Robustness helps sort explanatory dependencies, allocate evidential weight, and refine the model–world bridge, but does not itself confirm a fixed target. The result reorients RA from a confirmatory ideal to a constructive methodology in a complex, physiological science.
Khan, A. (2026). “Why Energy Balance Must Hold (and Why That Doesn’t Explain Obesity),” 51st Annual Meeting of the Society for Exact Philosophy (SEP). Vancouver, BC, Canada.
Abstract: Obesity science is haunted by a strange “must-hold” problem. Everyone agrees the energy balance principle has to hold, yet invoking it often sounds either like an uninformative slogan (“calories in, calories out”) or an empty truism. This paper explains that tension by showing why treating energy balance as a law-like generalization predictably backfires. If “energy intake” and “energy expenditure” are fixed by definition, the principle becomes unfalsifiable and uninformative; if they are treated as simple observables, the principle looks false in real-world settings. I argue that the way out is to stop asking what kind of law energy balance is, and instead ask what kind of modelling architecture makes it operative. On my proposal, energy balance is best understood as a property of a class of admissible models: models that explicitly represent energy stores and energy flows, constrain their trajectories with an accounting relation, and tie those variables to measurement and intervention practices. Seen this way, energy balance earns its “must-hold” status without pretending to be a complete explanation of obesity: it rules out incoherent stories, stabilizes intervention-oriented predictions in dynamic models, supplies a shared interface for mechanistic disagreements, and guides diagnosis when models and data conflict.
Khan, A. (2025). “Philosophical Perspectives on the History of Causal Modelling in Obesity Research,” Canadian Society for the History and Philosophy of Science (CSHPS). Toronto, ON, Canada.
Abstract: This talk examines obesity research as a historical case study for understanding causation in the special sciences. It explores the debate between energy balance and energy storage approaches to obesity by reconstructing historically significant episodes. I argue that an interventionist conception of causation clarifies a central theme in the field: researchers repeatedly treat manipulable variables as explanatorily privileged and view stable change under intervention as a mark of causal significance. However, this perspective is also incomplete. The obesity debate has consistently returned to questions that go beyond the identification of interventions. The dispute between energy balance and energy storage approaches, therefore, reflects a disagreement about the type of causal understanding obesity science should pursue. I conclude that this partial history motivates a pluralist framework for causal thinking in obesity research, while highlighting the substantial philosophical work still required to fully articulate such a framework.
Poster Presentations
Khan, A. (2026). “Robustness Analysis and the Energy Balance model of Obesity,” Rotman Graduate Student Conference 2026: Philosophical Issues in the Life Sciences. London, ON, Canada.
Abstract: This talk reassesses the evidential role of robustness analysis (RA) in model-based science through a case study of obesity modelling, focusing on Hall et al.’s (2020) Energy Balance Model (EBM). I argue that the standard RA taxonomies—within-framework, across-framework, and explanatory RA (ERA)—presuppose a strict model–target boundary and therefore cannot, on their own, underwrite world-directed confirmation in this domain. Within-framework RA remains diagnostic but intra-model; across-framework RA expands coverage but inherits incompatible idealizations; and ERA’s eliminative logic falters unless one assumes a contentious conjunctive adequacy bridge from model detections to worldly facts. Drawing on Hall et al.’s model, and the literature that they survey, I show that the explanandum “obesity” is iteratively re-specified alongside its model. I thus propose “iterative model–target co-construction” as the right way to understand this case. Robustness helps sort explanatory dependencies, allocate evidential weight, and refine the model–world bridge, but does not itself confirm a fixed target. The result reorients RA from a confirmatory ideal to a constructive methodology in a complex, physiological science.