It is commonly said that exchanging two identical particles is equivalent to rotating one of them by one revolution. The present note shows that, in this equivalence, both the exchange and the rotation are purely mathematical operations—symbolic manipulations of the wave function, made during the design phase of an experiment. Neither occurs in the physical spacetime in which the experiment runs. While in general, “exchange” and “rotation” refer to physical processes, in the context of this equivalence, they do not. The key observation is that either one can return the system to its original state perfectly—which is possible only for a mathematical operation, not for a real physical process. At best, it is an idealized physical process—the limiting case of a real one. However, in this equivalence, those two mathematical operations cannot even be considered as limiting cases, because their very definitions are purely mathematical. A physical process—real or idealized—is a function of time. Neither exchange (as particle label-swap) nor rotation (as frame coordinate transformation) has the time variable. The very meaning of “process” is the set of intermediate steps. Therefore: no time, no process. They are proposed mathematical properties of possible physical worlds. Mathematical properties—once well-defined or proved—are always true within their formal system. However, those properties can be inapplicable in some worlds. Therefore, it is possible to use physical experiment to test whether they apply in this physical reality. But they are still not physical processes within any particular spacetime. While Euclidean geometry is mathematical, the statement “this world follows Euclidean geometry if we define straight lines as paths of light” is physical, because it is possible to be falsified by experiment. Still, the geometric assertions never refer to any physical processes. As the operations are purely mathematical, the equivalence itself is likewise mathematical. Moreover, since either operation can exist without the other, the equivalence is not a logical necessity, but a convention.
It is commonly said that exchanging two identical particles is equivalent to rotating one of them by one revolution. The present note shows that, in this equivalence, both the exchange and the rotation are purely mathematical operations—symbolic manipulations of the wave function, made during the design phase of an experiment. Neither occurs in the physical spacetime in which the experiment runs. While in general, “exchange” and “rotation” refer to physical processes, in the context of this equivalence, they do not. The key observation is that either one can return the system to its original state perfectly—which is possible only for a mathematical operation, not for a real physical process. At best, it is an idealized physical process—the limiting case of a real one. However, in this equivalence, those two mathematical operations cannot even be considered as limiting cases, because their very definitions are purely mathematical. A physical process—real or idealized—is a function of time. Neither exchange (as particle label-swap) nor rotation (as frame coordinate transformation) has the time variable. The very meaning of “process” is the set of intermediate steps. Therefore: no time, no process. They are proposed mathematical properties of possible physical worlds. Mathematical properties—once well-defined or proved—are always true within their formal system. However, those properties can be inapplicable in some worlds. Therefore, it is possible to use physical experiment to test whether they apply in this physical reality. But they are still not physical processes within any particular spacetime. While Euclidean geometry is mathematical, the statement “this world follows Euclidean geometry if we define straight lines as paths of light” is physical, because it is possible to be falsified by experiment. Still, the geometric assertions never refer to any physical processes. As the operations are purely mathematical, the equivalence itself is likewise mathematical. Moreover, since either operation can exist without the other, the equivalence is not a logical necessity, but a convention.
Exchange forces lie at the base of some of the most striking explanatory successes of modern physics.An exchange interaction, for instance, is held to be responsible in some key sense for the chemical bond.An exchange force is the origin of ferromagnetism, providing the force that lines up individual spins into an overall pattern.In nuclear physics an exchange force is characteristic for the description of nucleon-nucleon forces.And, in perhaps the most far-reaching example, an exchange interaction is taken as definitional for the quantum field-theoretic description of nature, in which not only the fundamental forces (gravity, electromagnetism, the weak and strong nuclear forces), but also such non-fundamental ones as that operative between pairs of superconducting electrons in a lattice, are traced to the exchange of field quanta.In all these cases, moreover, exchange forces are a peculiarly quantum mechanical phenomenon, something characteristic of the new era that began in 1925 and continues into the present.And as a peculiarly quantum mechanical phenomenon, they are central to the new understanding of matter and force that was ushered in by the theory.For such a central concept, the historical origins of exchange forces are distinctly unclear.What was the path by which these diverse manifestations of exchange-in atomic physics, in condensed matter, in nuclear physics, in quantum field theory-came to go by the same name?Indeed, what is the common thread running through all these uses?The very notion of exchange is nebulous.Consider the following definition from a technical dictionary: exchange [QUANT MECH] (1) Operation of exchanging the space and spin coordinates in a Schrodinger--Pauli wave function representing two identical particles; this operation must leave the wave function unchanged, except possibly for a sign.(2) Process of exchanging a real or virtual particle between two other particles.'Here two distinct meanings of the term 'exchange' are on offer.The first notion
“Nothing changes” is used for three different operations in the treatment of identical particles. The operations are not equivalent, and the sentence changes meaning as it moves among them. Passive relabeling, A^'=P_12 A, changes only mathematical labels. In the standard quotient description, one physical situation is represented twice. The labels were never observables, and they supply no evidence that “electron A” and “electron B” are independently identifiable individuals. That does not forbid every metaphysics of primitive individuality. It makes such individuality empirically idle. Exchange symmetry is a different fact. For a state already in the fermionic subspace, P_12 ψ=-ψ. The isolated global sign is not itself a new pointer reading. Antisymmetry still constrains the allowed state space and has observable consequences, including exclusion. Active physical exchange is a process. A Hamiltonian can carry excitations along paths. Hong–Ou–Mandel and the Neder two-electron experiment show exchange-statistics interference in joint detection. Tschernig and collaborators directly measured the particle-exchange phase of photons by interfering a two-photon state with its physically permuted version. A braid is not a relabeling. In standard three-dimensional treatments the familiar sectors for observed elementary particles are bosonic and fermionic; parastatistics remains a formal complication. Anyonic statistics beyond +1/-1belong to braid structure in two-dimensional configuration spaces, typically effective quasiparticles, and do not reclassify ordinary electrons. The note draws two boundaries and refuses both overreaches. The same final invariant state does not imply the same physical process. A different process with no surviving record does not imply an additional present physical distinction. A difference of history counts as a present physical distinction only insofar as it remains written in the state, in a record, or in an observable relation. When no such mark survives, the final state does not contain an additional empirically discriminated distinction between the histories. This does not make the processes identical, and it does not make either process unoccur. This note introduces no modification of identical-particle quantum mechanics. Its claim is about inference. A statement valid for relabeling, for a symmetry transformation, or for physical transport cannot be carried from one operation to another without re-establishing what remains invariant. What this note is not It is not a correction of quantum mechanics. It is not a claim that electrons have hidden individual histories. It is not a claim that every metaphysics of primitive thisness is logically impossible. It is not a result about ordinary electrons in three dimensions ceasing to be fermions.
General remarks As already mentioned in the Introduction, the exact solution of the main equation of quantum mechanics – the Schrödinger equation – lies beyond the potentialities of modern mathematics and computer technology. But a number of important inferences about the behaviour, structure and properties of a given quantum-mechanical many-particle system can be drawn without solving this equation, just by examining its symmetry properties. Symmetry concepts come about in physics in two ways. First, since any physical process occurs in a real space, we have to make use of one or another coordinate system. The isotropy and homogeneity of space make physically meaningful only those mathematical relationships that remain unchanged under rotations of the axes of the coordinate system, and that impose fairly rigorous constraints on the possible physical laws. Second, every physical object and process features a symmetry which should be taken into account by the physical theory. A common idea underlying particular forms of symmetry is the invariance of a system under a certain set (group) of transformations. The normally considered forms of symmetry are rotational symmetry, which is based on the equivalence of all directions in space, and permutation symmetry, which is caused by identical particles. The operations of the geometrical symmetry group are responsible for appropriate conservation laws. So, the rotational symmetry of a closed system gives rise to the law of conservation of angular momentum. The group of rotations of a three-dimensional space stands apart in atomic spectroscopy. This is mostly due to the high accuracy of the central field approximation, on which the entire modern theory of complex atoms and ions is based.
We examine historic formulations of the spin-statistics theorem from a point of view that distinguishes between the observable consequences and the ``symmetrization postulate''. In particular, we make a critical analysis of concepts of particle identity, state distinguishability and permutation, and particle ``labels''. We discuss how to construct unique state vectors and the nature of the full state descriptions required for this -- in particular the elimination of arbitrary 2π rotations on fermion spin quantization frames and argue that the failure to do this renders the conventional symmetrization postulate (and previous ``proofs'' of it) at best {\em incomplete}. We discuss particle permutation in a general way for any pairs of particles, whether identical or not, and make an essential distinction between exchange and pure permutation. We prove a revised symmetrization postulate that allows us to construct state vectors that are naturally symmetric under pure permutation, {\em for any spin}. The significance of particle labels (which, in the exchange operation, are not permuted along with other variables) is then that they stand in for any asymmetry (order dependence) that is present in the full state descriptions necessary for unique state vectors but not explicit in the regular state variables. {\em The exchange operation is then the physical transformation that reverses any asymmetry implicit in the labels}. We point out a previously unremarked geometrical asymmetry between all pairs of particles that is present whenever we choose a common frame of reference. We compute the exchange phase for various state vectors using different spin quantization frames, and prove the Pauli Exclusion Principle and its generalization to arbitrary spin.
We discuss the conditions under which identical particles may yet be distinguishable and the relationship between particle permutation and exchange. We show that we can always define permutation-symmetric state vectors. When the particles are completely indistinguishable, then exchange is equivalent to permutation and therefore the exchange eigenvalue for such permutation-symmetric state vectors is always +1. Exchange asymmetry arises when the particles are physically distinguishable, even though otherwise identical, and can be computed from the transformations that arise when the distinguishing features are reversed. There is a fundamental spatial asymmetry between the relative orientations of any two vectors in a common frame of reference that persists even in the limit that the vectors coincide. For a pair of particles this asymmetry between their spin quantization frames renders them distinguishable even when otherwise identical. In the conventional construction, this distinction is not properly accounted for. Particle exchange is then equivalent to reversing this relative orientation --- which requires a relative rotation by 2pi on the spin quantization frame of one particle with respect to the other, thus resulting in the conventional exchange phase.