Volume 27 · Part Six · Chapter 19 of 19

Keeping the Full Description in View

The practice, stated as a procedure.

The procedure

State the full object first. Before borrowing, write down what the borrowed structure is a piece of — for this volume, the metric and the Riemann tensor of full General Relativity, with no linearisation and no effective medium.

Name the proxy at adoption. Not in a footnote and not on review: in the sentence that introduces it. 'We treat the field as a medium with parameters ε_g and μ_g; this is valid to first order in h.'

Write the band as a number, not a word. 'Weak field' is not a band. A stated bound on the neglected term is.

Attach a falsifier. If no measurement could distinguish the proxy from the full object, the proxy is not doing physics, and should be labelled as an aid to thought.

Re-state on every reuse. Labels decay across citations faster than results do. The re-statement is the maintenance.

Why it cannot be automated

Each step requires a judgement about purpose — what the borrowing is for — and purpose is exactly what an optimisation process does not represent. A checker can verify that a status label is present. It cannot verify that the label is the right one, because that requires knowing what the work is trying to be true about. This is the sense in which the requirement is standing: it is not a gate to pass but a practice to keep up, and it is the part of the work that stays human for structural reasons rather than sentimental ones.

The volume ends where it began. The borrowings are good. Riemann's geometry is generous and it lends well. The discipline is only this: return what you borrowed, in the condition you found it, and never mistake the loan for the estate.

Equations borrowed

  • Nothing. This chapter states the volume's own rule.

Validity band

Applies to every borrowing in this volume and, the author would argue, to the corpus.

Falsifier

A body of work that abandoned the practice and produced durable results anyway would weaken the claim that the requirement is necessary rather than merely tidy.

Where this chapter is weakest

The chapter prescribes without measuring. It offers no evidence that groups following this procedure produce better physics than groups that do not, and it should not pretend the discipline is more than a well-motivated proposal.

The volume-wide audit of these weak points is collected in Where This Volume Is Weak.