The Rock and the Plane
Economist Robert P. Murphy offers a simple thought experiment: a rock and a plane are both flying through the air. How should we understand their trajectories? This is not a question of physics alone but of the lens we apply to explain motion. The rock invites one kind of analysis, while the plane demands another.
The Objective Prior
When we watch the rock fly, we use what can be called an objective prior. We assume the system follows mechanical laws: gravity, momentum, drag. The explanatory framework begins from atoms and forces. This is materialist reductionism at work. We can write equations, run simulations, and with enough precision predict the exact arc the rock will make. Atoms move according to causes that unfold deterministically, and this framework fits the case perfectly. When we study such systems, it is almost artificial to invoke 'intentions' or 'decisions.'
The Subjective Prior
The airplane complicates things. Of course it also obeys aerodynamics and Newtonian mechanics, but when we ask where the plane is going, the relevant variables shift. Its pilots have plans, its passengers have destinations; the route depends on choices encoded into systems of action. Here we use what can be called the subjective prior: we begin from situated agents making decisions from within constraints. We learn more by asking 'why is the pilot banking left?' than by attempting to track every atom in the wings. Reductionism still holds, but it is inefficient compared to the lens of chosen action. In this frame, even the autopilot’s code is a form of embedded plan, a frozen decision-making structure.
Determinism and Stochasticity
Neither prior collapses into randomness. At bottom, both the rock and the plane follow deterministic causal chains. If we had perfect access to conditions, we could trace exact outcomes. But as observers, we often lack that access. What shows up as probabilistic models of flight paths is not evidence of inherent unpredictability but of our inability to observe every decision, constraint, or physical parameter. Stochasticity is a feature of our models, not of the systems themselves. The rock’s arc looks simple enough that objective priors suffice; the plane’s path involves enough hidden decisions that subjective priors are indispensable.
Choosing the Right Lens
The rock and the plane show us why we must choose carefully between priors. To collapse everything to physical matter obscures the meaning that drives events in human systems. To collapse everything to subjective choice denies the lawful order present in physical mechanisms. The real task is knowing which trajectory is better understood atomically and which is better understood through situated action. Navigating this distinction is the same task transformers face when they compute meaning through geometry. As described in Welcome to Meaningspace, the structure of meaning is not reducible to information alone but depends on perspective and situatedness. Our theories must honor that difference.