Explaining what happened earlier is not the same task as explaining what something depends on. A cause, a definition, a constitutive relation and a proposed metaphysical ground can all answer why questions while doing different work.
Different questions behind why
Aristotle's Physics, book II, chapter 3, distinguishes several senses of cause: material, form, the source of change and that for the sake of which something occurs. His classification reminds a modern reader that why need not request only an earlier initiating event.
That historical classification should not be identified wholesale with contemporary grounding theories. It provides an entry point into distinctions that must be reconstructed for the particular argument being read. Different uses of cause may be legitimate without being interchangeable.
Ask why a bridge carries a load. The answer might describe its construction history, the arrangement of members, the properties of steel or the engineer's design objective. None of these answers automatically replaces the others. Each selects an explanatory target.
Dependence without a delay
A simple bookkeeping example separates dependence from temporal succession. Let a report define total T as a+b. For entries a=3 and b=5, T=8. The total depends on the entries under this definition; the defining relation does not say that nature waits a measurable interval before eight becomes true.
A computer implementing the addition may take time. That processing history is another description, involving hardware and operations. Confusing the definition with its implementation would turn every arithmetic relation into a claim about a particular processor.
Even a simultaneous equation does not, by itself, establish a metaphysical direction of dependence. T=a+b can be rearranged as a=T−b. Algebraic equivalence preserves the numerical constraint; it does not decide whether totals depend on parts, parts on totals or both descriptions express some further structure.
A constraint and a mechanism
Consider the constraint x+y=10. It permits many pairs: (0,10), (4,6), (10,0) and more if real values are allowed. The constraint alone supplies no rate of motion between them. One may stipulate x changes by one per step and y decreases by one, or instead specify that both remain fixed.
Both dynamics preserve the constraint. They disagree about future observations. Deriving a relation between observables is therefore different from deriving the mechanism that carries a system through allowed states.
This example is deliberately elementary. It identifies a common gap in ambitious explanations: a successful structural identity is presented as if it already determines an evolution law. Filling that gap requires an additional argument, not a more emphatic restatement of the identity.
The direction of an explanation
An explanation can also fail by moving in a circle. Suppose a length scale is selected to reproduce a measured frequency, and that frequency is then announced as derived from the selected length. The calculation may be internally consistent, but it has not supplied an independent prediction.
Draw a dependency list: assumptions, definitions, calibration data, intermediate results and predicted outputs. Mark which steps are deductions and which are identifications with experiment. If an input already contains the target observation, the result should be described as reconstruction or fitting at that point.
A metaphysical grounding proposal has related but distinct obligations. It must state what depends on what, why the dependence has that direction and how the proposal improves understanding. These are philosophical tasks; they should not be confused with a numerical fit or a laboratory intervention.
Generation as a grounding proposal
SMM/DTT proposes one source and distinguishes generative inner roles from accessible outer records. The book's reconstruction requirements identify the additional content needed for physics: histories, completion, dynamics, probabilities, comparisons, clocks and observables.
One source is an organizing commitment. An explanatory physical realization must show how that commitment restricts the permitted structures. If the desired geometry, phase rule and clock law can be chosen independently, the work of connecting them has not yet been completed.
The distinction also clarifies debates about creation. A temporal beginning and continuing dependence are different questions. Explaining one need not settle the other, and a diagram that places a source before an output should not quietly turn every dependence into an earlier event.
The resulting discipline is constructive: identify the kind of explanation being offered, state its premises and show what follows that could not simply be chosen at will.
Sources and further reading
- Aristotle, Physics II.3, translated by R. P. Hardie and R. K. Gaye.
- Mohamed Haj Yousef, What a reconstruction must contain and The place of philosophy.