This text critiques the nature and role of scientific models, highlighting the views of statistician George Box and philosopher Hilary Lawson. It explains that Box is known for the idea that all models are flawed but some are still useful, a concept he expressed by stating that scientists must be aware of significant errors in their models. The piece then presents Lawson’s more radical perspective, suggesting that models don't require truth to be effective and are merely pragmatic tools for interacting with an inherently unknowable reality. The text uses the historical example of celestial spheres to illustrate how operational utility can trump factual accuracy, concluding that models are ultimately useful approximations or "lies that work" rather than representations of absolute truth.https://philosophics.blog/2025/04/25/unwilling-steelman-part-iv/