Change and persistence concern different features of a history. A quantity can remain constant while the state changes; an organization can persist while its components change. Neither constancy nor resemblance alone establishes which history is being continued.
What exactly has remained?
When we say that something persists, we may mean its material, its organization, its function or its connection to an earlier preparation. These criteria often agree in ordinary life, which makes their differences easy to overlook. They become visible when components are replaced, records are copied or a system returns to a familiar configuration.
The first useful step is to state the candidate invariant: a quantity or relation that remains unchanged under a specified transformation. The next step is to ask how much that invariant identifies. A conserved quantity constrains a history, but it rarely gives a complete description of that history.
Constant energy, changing phase
Take an ideal harmonic oscillator in normalized units, with position x, momentum p and equations dx/dt = p, dp/dt = −x. Its energy is E = (x² + p²)/2. Differentiating gives dE/dt = xp − px = 0. The constancy follows from the stated dynamics.
Starting at (1, 0), the oscillator reaches (0, −1) after a quarter period, (−1, 0) after a half period and (0, 1) after three quarters. Every state has energy 1/2. Yet the position and direction of motion differ. Knowing the conserved energy does not tell us which of these states is present.
The explicit solution, x = cos t and p = −sin t, also shows what additional information matters: the phase relative to an agreed preparation or clock. Energy fixes the size of the orbit. Phase locates the state along it. This familiar mechanical example separates a continuing constraint from a complete specification.
A return does not erase the elapsed history
After one period, the ideal oscillator returns to its initial pair of values. If another register has counted the completed cycles, the combined oscillator-and-counter system has not returned. What counts as a repetition depends on which system and which records are included.
Conversely, two separately prepared oscillators can have identical energies and phases at a chosen time. Agreement of their states does not make them the same preparation. To establish a claim about origin, we need a connection to that origin: a record, a tracked interaction or a specified preparation procedure.
This is why identifying an object and identifying a state are related but different tasks. The same object can occupy different states. Distinct objects can occupy states that agree in the variables being measured. A physical account should say which kind of identity its calculation addresses.
Persistence can survive the loss of an invariant
Add damping: dp/dt = −x − γp, with positive γ. The same energy expression now obeys dE/dt = −γp². Energy of the oscillator decreases as its interaction with the surroundings becomes relevant. That change does not by itself imply that the oscillator has ceased to be the system under investigation.
Many starting states approach the same equilibrium. Their eventual resemblance therefore cannot reconstruct their earlier differences without further records. A stable end state and a retained history are different achievements. This observation connects persistence to the information needed for prediction, as well as to what can be inferred about the past.
Renewal needs relations across updates
The expanded Duality of Time book treats continuing renewal as compatible with carried correlations. Its section on renewal, persistence and a continuous limit examines how a later record can depend on earlier records even when the description uses successive completions. The proposed unity of generation does not remove the need for a sufficiently rich state.
For SMM/DTT, a concrete challenge follows: specify the update that preserves the distinctions needed by later experiments. It is not enough to describe a sequence of similar configurations. A model must explain their connection, including which phase, memory or other information survives. The oscillator calculations above illustrate that requirement; they do not derive a unique mechanism of renewed creation.
Choose the persistence question first
Different investigations can legitimately follow different continuities. An energy audit tracks transfers. A synchronization experiment tracks relative phase. An investigation of provenance tracks preparation and interactions. The word “same” becomes useful when attached to one of these explicit questions, with enough evidence to support the corresponding relationship.
Sources and further reading
- Feynman, Leighton and Sands, The Feynman Lectures on Physics, I.21: The Harmonic Oscillator; the normalized calculations here are worked explicitly.
- Mohamed Haj Yousef, Renewal, persistence, and a continuous limit.
- Continue with The identity of an object through change and Cycles and renewal.