Source
Rovelli (1996), LIT-1635, arXiv quant-ph/9609002v2; read in NOTE-1227. Page numbers are the preprint’s.
What was actually shown
What is relational. “The quantum state of a system is always a state of that system with respect to a certain other system” and “In quantum mechanics all physical variables are relational, as is velocity” (§II.C, p. 6). “There is neither an absolute state of the system, nor absolute properties that the system has at a certain time” (p. 7).
What is not. The reconstruction begins: “Assume that the world can be decomposed (possibly in a variety of ways) in a collection of systems, each of which can be equivalently considered as an observing system or as an observed system” (§III.A, p. 10). A system “is characterized by the family of yes/no questions that can be asked to it”, W(S), whose structure gives “the general kinematical features of S” (pp. 10–11); by Postulate 1 “each system is characterized by a number N”, its information capacity (p. 11); and W(S) “includes all Heisenberg observables at all times”, so it fixes the Hamiltonian (§III.D, p. 15). The summary repeats it: “A physical system is characterized by the structure on the set W(S) of questions that can be asked to the system” (§IV.A, p. 16). None of these is relative to an observer. A state exists “only if some system O … has interacted with S” (p. 6): the interaction is between systems already there.
What is absent. The 1996 text has no ontology of events or facts. Its primitives are systems, questions and information (correlation); “the state of the system” is replaced by “the information that a system has about another system” (p. 11). The “sparse ontology of (relational) quantum events” that Faye and Jaksland quote (LIT-1672, §5) is from later writing.
The reading could have come out otherwise: the paper could have made a system’s identity relative too, or taken interactions as primitive with systems derived from them. It does neither.
What this does not say
- That the priority is argued. It is an assumption (“Assume”), made for the reconstruction, and the paper declines to discuss it philosophically (footnote ‡‡, p. 19).
- That the systems are uniquely given. The decomposition is “possibly in a variety of ways”, and how different decompositions relate is not discussed.
- That a system is individuated. W(S) and N fix what kind of system it is (a lattice of subspaces of a k-dimensional Hilbert space). Nothing in the paper says what makes two systems of one kind two, which is CLAIM-136’s question, not answered here.
- That RQM in every later form keeps this structure. That is for the later texts (LIT-177 and those Faye and Jaksland quote).
- That the questions constitute the system. “Characterized” is Rovelli’s word; CLAIM-477’s stronger “partly fixed by the constraints” is not his.