The Sock Problem

A three-year-old can discover that socks followed by shoes produces a dressed foot while shoes followed by socks does not. The joke is algebraic: order matters.

Let $A$ mean “put on socks” and $B$ mean “put on shoes.” On a simplified state space, $BA$ is defined and reaches the intended state; $AB$ may be blocked or lead somewhere else. These are partial, state-changing operations, not group elements: they need not be invertible or defined on every state. “Non-abelian pre-schooler” is therefore analogy, not classification.

The same idea appears more cleanly in rotations. Rotate a book around one axis and then another; reverse the order and its final orientation generally changes. The operations are well-defined rotations, and in three dimensions they do not generally commute.

What the Developmental Studies Show

Piaget’s concrete-operational stage concerns a cluster of abilities including logical operations and reversibility, not the claim that younger children know nothing about event order. Later work shows that infants and young children can encode and reproduce causal or temporal sequences under specified tasks.

Bauer and Mandler studied recall of multi-step events in children between one and two years old. Klemfuss and colleagues found that young children’s answers to sequencing questions were strongly affected by encoding order. Those results concern memory, causal structure, question form, and temporal order.

They do not demonstrate that toddlers represent binary operations, test $AB=BA$, understand inverses, or anticipate Hilbert-space commutators. Nor does successful sock order refute Piaget’s account of formal reversibility. The pedagogical point is modest: learners have everyday experiences in which order changes an outcome, and those experiences can motivate an algebra lesson.

Matrices and Quaternions

Two operations commute when $AB=BA$. Matrix multiplication supplies a concrete counterexample:

$$A=\begin{pmatrix}1&2\\3&4\end{pmatrix},\qquad B=\begin{pmatrix}0&1\\1&0\end{pmatrix}.$$

Then

$$AB=\begin{pmatrix}2&1\\4&3\end{pmatrix},\qquad BA=\begin{pmatrix}3&4\\1&2\end{pmatrix}.$$

The products differ. Arthur Cayley’s 1858 memoir helped establish matrix algebra as a general theory, although matrices and related arrays had earlier uses.

Hamilton’s quaternions give another famous relation:

$$i^2=j^2=k^2=ijk=-1,$$

from which $ij=k$ while $ji=-k$. Their noncommutativity is useful in describing three-dimensional rotations. It is not the same algebra as the cyclic transposition action used in the Messiaen article; the previous cross-reference confused distinct structures merely because both use group language.

Quantum Operators

In classical Hamiltonian mechanics, position and momentum are coordinates on phase space and their Poisson bracket satisfies $\{x,p\}=1$. In matrix and operator quantum mechanics, the canonical relation is

$$[\hat x,\hat p]=\hat x\hat p-\hat p\hat x=i\hbar I.$$

Heisenberg’s 1925 reformulation and Born and Jordan’s matrix treatment are central steps in that history. The products in the commutator are compositions of linear operators. They are not literally “measure position, then measure momentum” as two laboratory instructions; sequential quantum measurement requires an additional measurement model.

For a normalised state in the relevant operator domains, Robertson’s inequality states

$$\Delta A\,\Delta B\ge \frac12\left|\langle[\hat A,\hat B]\rangle\right|.$$

With the canonical commutator this gives

$$\Delta x\,\Delta p\ge\frac{\hbar}{2}.$$

Noncommutativity is essential to this bound, but the relationship is not simply “every noncommuting pair has a fixed nonzero uncertainty.” The expectation of a commutator can vanish in some states, domains matter for unbounded operators, and the Schrödinger relation includes a covariance term beyond Robertson’s commutator bound.

Angular-momentum components satisfy

$$[\hat L_x,\hat L_y]=i\hbar\hat L_z$$

and cyclic permutations. The resulting lower bound is state-dependent, for example $\Delta L_x\Delta L_y\ge(\hbar/2)|\langle L_z\rangle|$. This is more precise than saying no two components can ever be simultaneously sharp without specifying the state and observable pair.

The sock example and the quantum relation therefore do not have the “same reason.” They share a formal warning: do not exchange operation order until the algebra says you may.

Noncommutative Geometry

Ordinary spaces can be studied through commutative algebras of functions. Noncommutative geometry generalises this strategy to noncommutative operator algebras. A spectral triple $(\mathcal A,\mathcal H,D)$ consists, roughly, of an algebra represented on a Hilbert space and an operator carrying geometric information, subject to technical conditions omitted here.

Connes and Chamseddine developed spectral-action models whose chosen almost-commutative geometry yields a form of the Standard Model coupled to gravity. The result depends on the algebra, representation, cutoff action, symmetry breaking, and parameter assumptions. It is a research programme, not a derivation showing that “space itself” is empirically noncommutative or that all Standard Model and gravitational physics follows without input.

That is already enough range for one idea: from finite matrices to quantum observables to a programme in geometry. The preschool anecdote is the door, not the evidence.

The Lesson

Students accustomed to scalar arithmetic may overgeneralise commutativity to matrices. An instructor can answer with a book rotation or socks and shoes, provided the analogy’s missing properties are stated. Everyday action order is familiar. Abstract closure, inverses, linearity, domains, states, and uncertainty bounds are new work.

The universe is not uniformly non-abelian, either. Integer addition commutes; spatial rotations generally do not; particular quantum observables may or may not commute. Algebra tells us which situation we are in.

The child knew which garment came first. Physics begins after we stop pretending that this alone proves the commutator.

Mathematics, physics, and developmental literature checked through 2026-07-11.

References

  • Bauer, P. J., & Mandler, J. M. (1989). One thing follows another: Effects of temporal structure on 1- to 2-year-olds’ recall of events. Developmental Psychology, 25, 197–206.

  • Klemfuss, J. Z., McWilliams, K., Henderson, H. M., Olaguez, A. P., & Lyon, T. D. (2020). Order of encoding predicts young children’s responses to sequencing questions. Cognitive Development, 55, 100927. https://doi.org/10.1016/j.cogdev.2020.100927

  • Cayley, A. (1858). A memoir on the theory of matrices. Philosophical Transactions of the Royal Society of London, 148, 17–37. https://doi.org/10.1098/rstl.1858.0002

  • Born, M., & Jordan, P. (1925). Zur Quantenmechanik. Zeitschrift für Physik, 34, 858–888.

  • Robertson, H. P. (1929). The uncertainty principle. Physical Review, 34, 163–164. https://doi.org/10.1103/PhysRev.34.163

  • Connes, A. (1994). Noncommutative Geometry. Academic Press.

  • Chamseddine, A. H., & Connes, A. (1996). Universal formula for noncommutative geometry actions. Physical Review Letters, 77, 4868–4871. https://doi.org/10.1103/PhysRevLett.77.4868

Changelog

  • 2026-02-03: Corrected the Klemfuss et al. age range and encoding-order result.
  • 2026-07-11: Recast socks and shoes as a partial-operation analogy rather than the cause of quantum uncertainty; bounded developmental evidence; corrected Robertson/angular-momentum scope; and narrowed noncommutative-geometry claims.