Why Fields Replace Wavefunctions

Relativistic quantum field theory replaces a fixed collection of particle coordinates as the universal starting point with quantum fields and their local algebra. Quantum states, superposition, and probability amplitudes remain. A one-particle wavefunction can reappear as a matrix element of a field operator, while a general state need not lie in any one particle-number sector. The central transition is therefore a change in the dynamical objects and their organization, not the abandonment of wavefunctions or the Born rule.

Required background. Why Fixed Particle Number Fails identifies the dynamical limitation; Why Fock Space Is Necessary supplies variable-number state spaces; Locality and Causality Warnings separates correlation from causal response.

Helpful background. Dirac Equation as Bridge lists the wave-mechanical ingredients inherited by a quantized spinor theory.

A wave equation leaves the quantum structure unspecified

An equation such as $(\Box+m^2)f=0$, written here in natural units $\hbar=c=1$, fixes the evolution of a numerical mode function. The same differential operator can act on a classical field, on a suitable one-particle amplitude, or on a free operator field. The differential equation alone does not decide which interpretation is intended.

Quantization requires additional structure:

• a field algebra, including the relevant commutators or anticommutators;
• a representation and quantum state, such as a specified vacuum or density operator;
• dynamics and an interaction prescription;
• a definition of observables and of the products or correlations used to predict their statistics.

Boundary conditions and state preparation also matter. A retarded differential inverse and a vacuum time-ordered two-point function can be closely related without being the same object. The later propagator-to-correlator bridge makes that distinction explicit.

In particular, ordinary four-component Dirac amplitudes do not acquire fermionic statistics merely because they transform as spinors. The relativistic field construction must supply an operator algebra compatible with its physical assumptions. The free spinor-field calculation checks that compatibility; a general spin–statistics theorem needs its own hypotheses.

States, fields, and numerical amplitudes

Keep the following objects separate:

Object	Meaning	What its arguments label
A state vector $\lvert\Psi\rangle$ or density operator $\rho$	The quantum preparation	No spacetime point is required to define the abstract state
A numerical mode $f(x)$ or $u_s(p)$	A solution or basis coefficient	Spacetime or on-shell momentum labels
An operator field $\widehat\phi(x)$	A spacetime-indexed operator-valued distribution	The location of a field insertion
A wavefunctional $\Psi[\varphi]$	A state in a field-configuration representation	An entire classical configuration $\varphi(\mathbf x)$

For a free scalar one-particle state $\lvert f\rangle$, an amplitude of the form

$$F_f(x)=\langle0|\widehat\phi(x)|f\rangle$$

obeys the free wave equation because the operator does. Its normalization and interpretation are developed in Fock Space to Quantum Fields. It is a matrix element between two states, rather than the operator $\widehat\phi$ itself. Its squared magnitude is not automatically a normalized local relativistic position probability.

An expectation value $\langle\Psi|\widehat\phi(x)|\Psi\rangle$ is another numerical field. It usually contains only a small part of the information in the state. A free vacuum has vanishing one-point function but nonzero two-point correlations. Many other states can share that same zero mean.

A regulated wavefunctional is still quantum mechanics

To make the configuration language concrete, keep a finite set of real bosonic normal-mode coordinates $q_1,\ldots,q_N$. A state has an ordinary coordinate wavefunction

$$\Psi(q_1,\ldots,q_N,t),\qquad \int d^Nq\,|\Psi(q_1,\ldots,q_N,t)|^2=1.$$

The coordinate operators multiply this function, and their conjugate momenta act as $-i\hbar\,\partial/\partial q_j$. The Schrödinger equation still evolves the state. In a field limit, the mode configuration corresponds to a spatial field configuration, and the notation $\Psi[\varphi,t]$ records this enlarged list of arguments. Defining its continuum measure and domain requires more than replacing a finite product by a formal symbol.

The coordinate $q_j$ is the amplitude of a field mode, not the position of the $j$th particle. Its oscillator excitation number can be interpreted as a mode occupation in the free theory. The oscillator bridge derives that mode construction; it is why field coordinates and particle occupations can describe the same regulated free state in different bases.

This bosonic coordinate example does not prescribe ordinary commuting field coordinates for fermionic functional integration. Fermion path integrals use Grassmann variables, which must be distinguished from both operator fields and numerical external spinors.

Fock space is not yet a local field theory

The Fock construction organizes particle number and exchange statistics. It includes a distinguished empty zero-particle vector, but does not alone identify that vector with the physical vacuum of chosen dynamics. Nor does it choose the Hamiltonian, a local field operator, or a Lorentz transformation law. Those ingredients must be specified and checked. For free relativistic fields, the relative weights of positive- and negative-frequency modes are consequential for both normalization and locality.

Locality is a condition on the field or observable algebra at spacelike separation. For fermionic fields one uses the corresponding graded relations; physical local observables commute when separated spacelike under the usual relativistic locality assumptions. This permits nonzero spacelike state correlations while excluding their interpretation as a controllable superluminal signal.

Fields at a point are generally distributions. For example, a real scalar field is used in a smeared form

$$\widehat\phi(g)=\int d^4x\,g(x)\widehat\phi(x)$$

with a suitable smooth test function. Real $g$ gives a Hermitian smeared field on an appropriate domain in the free scalar theory. Products at the same point, such as an interacting charge current, require a definition or renormalization prescription. A charged spinor field is not itself a Hermitian observable; in a gauge theory physical observables must also respect the gauge constraints.

The language also extends beyond relativistic systems. A number-conserving nonrelativistic many-body theory can be written as a genuine quantum field theory. The many-body route to fields explains that construction. Variable particle number is an important motivation here, but is not a universal definition of every field theory.

What survives in a particle description

Stable, well-separated incoming and outgoing particles can remain useful in an interacting theory. Their wave packets and scattering amplitudes recover much of the familiar particle language. In other settings the spectrum, correlations, and local response are more useful than a fixed-particle wavefunction.

This does not guarantee that every interacting infinite-volume state is a finite-particle free Fock vector, or that every theory has a simple particle-number observable. Vacuum choice, infrared behavior, and the existence of asymptotic particles are additional questions. The LSZ owner states the stable-pole boundary of its scattering construction.

The following pages make the transition operational: build fields from oscillator modes, relate field matrix elements to particle amplitudes, establish the free fermionic locality check, and then distinguish correlations, sources, scattering, and gauge dynamics. Each step adds information that the original one-particle differential equation did not contain.

Exercises

Equal means, different states. For one regulated oscillator mode with mass parameter one, $q=\sqrt{\hbar/(2\omega)}(a+a^\dagger)$. Compare $\langle q\rangle$ and $\langle q^2\rangle$ in the vacuum and the one-quantum number state.

Solution

Both means vanish because $a$ and $a^\dagger$ change the occupation. The variances are $\hbar/(2\omega)$ and $3\hbar/(2\omega)$. Thus the mean field alone cannot distinguish these two states, while a two-point measurement can.

Which space is being described? A scalar state has a configuration wavefunction $\Psi(q_1,q_2)$ for two retained normal modes. Does this necessarily describe exactly two particles?

Solution

No. There are two oscillator coordinates, each with its own occupation basis. A superposition of their number states can contain different total occupations. The number of configuration coordinates counts modes, not particle number.

What a mode equation omits. Two constructions use the same classical Klein–Gordon solutions. What additional data must agree before their quantum correlations can be identified?

Solution

They need compatible field normalization and algebra, the same state or vacuum prescription, and the same ordering and boundary conditions for the requested correlation. An interacting comparison also needs the same dynamics and operator-product definitions. Equality of the classical mode equation is insufficient.

References

• Schwartz, Matthew D. Quantum Field Theory and the Standard Model. Cambridge University Press (2014). doi:10.1017/9781139540940. States, field operators, local interactions, and scattering.
• Thaller, Bernd. The Dirac Equation. Springer (1992). doi:10.1007/978-3-662-02753-0. The operator and one-particle framework inherited by relativistic field constructions.
• Tong, David. Quantum Field Theory. University of Cambridge lecture notes (2006), sections 2 and 5. Free fields; quantizing the Dirac field. Canonical field quantization and its state-space interpretation.