Thermodynamics is Not Subjective, but Relational

A persistent concern in the foundations of thermodynamics is that its reliance on epistemic elements—agents, probabilities, and descriptive choices—renders it subjective. There are arguments to the effect that objective quantum mechanical probabilities can address this. We argue that such claims misdiagnose the sources of subjectivity in thermodynamics, which actually stem from its relational character. We illustrate this via the passivity framework—a rigorous foundation for the Gibbs state—bypassing probabilistic worries. We further show that the objectivity-subjectivity binary is too coarse to capture the conceptual landscape, for thermodynamic concepts are better understood through a richer taxonomy including the epistemic, relational, and means-relative. (Note: Conditionally accepted for publication in Philosophy of Science, PSA 2026 Conference Proceedings).