Idealization in Science

Core Idea

Scientific models are typically idealizations: they ignore complicating factors, assume unrealistic conditions, and employ false simplifications. Physics assumes frictionless surfaces; biology assumes infinite populations; economics assumes perfectly rational agents. Idealizations enable tractable analysis and explanation. However, they raise philosophical questions: How do false idealizations provide true explanations? When is an idealization legitimate versus misleading? How do we apply idealized theories to real systems?

Explainer

From your study of scientific models, you know that models are not literal descriptions of the world — they are purpose-built representations that highlight some features while ignoring others. The concept of idealization sharpens this: many models do not just omit details, they make claims that are *strictly false*. A frictionless plane does not exist. An ideal gas with no intermolecular forces does not exist. An infinite population with perfectly equal reproductive rates does not exist. Yet physics, chemistry, and biology routinely build theories on these fictions, and those theories make accurate predictions. This is the idealization puzzle: how can false assumptions generate true predictions?

One influential distinction is between Galilean idealization and Aristotelian abstraction. Aristotelian abstraction removes accidental features to reveal a simpler underlying truth — like abstracting from this specific triangle to properties that hold of all triangles. Galilean idealization, by contrast, deliberately introduces *false* assumptions to render systems tractable. Newton's *Principia* assumes that planetary bodies have their mass concentrated at a point; this is false for real planets, but it gives excellent predictions because the deviation from point-mass behavior is negligibly small. The key question is when such false assumptions are *legitimate* — when the conclusions you draw from the idealized model accurately describe the real system despite the false assumption.

The de-idealization problem asks: if we know the model is false, how do we correct it? Sometimes the idealization is a controlled approximation — we can add corrections and show that the idealized result is the limit as those corrections vanish. Other times, the idealization plays a deeper explanatory role: the infinite-population assumption in population genetics does not just simplify computation — it *generates* the Hardy-Weinberg equilibrium, which is used as a null model to detect selection in real finite populations. Here the false assumption is doing *explanatory* work, not just computational work.

A philosophically important case is infinite idealization: some explanations seem to require taking a system to an infinite limit (infinite population, infinite system size, thermodynamic limit). The puzzle is that real systems are never infinite, yet the explanatory concepts — phase transitions, spontaneous symmetry breaking — only appear at the infinite limit. This raises questions about whether such explanations are truly mechanistic or whether they are something else: structural, mathematical, emergent. Your study of laws of nature is relevant here: idealizations often work by identifying a true law governing an idealized system, then arguing that the real system closely approximates the ideal. Understanding when that argument succeeds and when it misleads is a central task in the philosophy of science.