Bridges to QFT and Statistical Field Theory

Field theory provides several useful languages for many-body physics. An operator field can exactly rewrite a particle model. A coherent-state field is an integration variable. An order parameter or hydrodynamic field usually describes selected collective behavior at long distances or times. These constructions have different meanings and assumptions.

This chapter helps you choose the formulation appropriate to a partition function, correlation, transition, or real-time problem. Begin with Why Many-Body QM Leads to QFT for the conceptual connections.

Required background. Many-Body Hilbert Spaces and Operators supplies sectors, modes, field operators, and operator Hamiltonians. Correlation Functions and Linear Response supplies sources, correlators, analytic structure, response, and observable conventions.

Helpful background. Emergence and Effective Degrees of Freedom prepares matching and scale-dependent variables; Finite-Temperature Methods prepares imaginary time; Phases, Order, and Criticality prepares order parameters and RG questions; Nonequilibrium Many-Body Dynamics prepares initial states and real-time protocols.

Separate an exact change of language from an effective theory

For complete orthonormal one-particle modes $\phi_r(\mathbf x)$, the expansion

$$\hat\psi(\mathbf x)=\sum_r\phi_r(\mathbf x)a_r$$

expresses the same annihilation operator in position language. It does not by itself discard states or introduce a low-energy approximation. A finite mode cutoff, by contrast, projects onto a restricted one-particle space and changes the spatial field algebra accordingly.

A properly regulated path integral can likewise represent the same quantum problem. Introducing an auxiliary field by a normalized Gaussian identity can be exact. Selecting one saddle point or dropping fluctuation terms is an additional approximation. Keep those steps separate so that an exact algebraic identity is not mistaken for a guarantee about the final approximation.

An effective field theory goes further: it retains variables suited to a scale window and matches their couplings and observables to the microscopic description or experiment. Specify the retained symmetries, slow modes, regulator, boundary conditions, and terms omitted at the chosen order. A hydrodynamic theory also needs a justified closure for its slow variables; conservation laws alone do not determine all constitutive relations.

For thermal questions, choose a Euclidean trace with its thermal boundary conditions. For an initial-value problem, preserve the initial density matrix and real-time ordering. Relativistic covariance and microcausality are extra physical principles; neither follows merely from writing a nonrelativistic field operator.

Choose a field-theory route

Choose the representation that matches the observable and time dependence.

1. Choose the conceptual trunk. Start with Why Many-Body QM Leads to QFT to distinguish operator fields, collective fields, effective theories, and relativistic QFT.
2. Take the operator-field route. Field operators, Fock sectors, and symmetries lead to Nonrelativistic Field Theory from Many-Body QM. Green functions then test the observable content. The field rewrite can be exact before any low-energy approximation.
3. Take a functional-integral route. Path Integrals for Many-Body Systems compares worldline, local-basis, and field representations. Coherent-State Path Integrals specializes bosonic, fermionic, and spin variables, overlaps, boundary data, Berry terms, and normal ordering.
4. Decouple an interaction only when needed. Hubbard–Stratonovich Transformation Preview separates exact auxiliary-field identities and exact fermion integration, when available, from incomplete channel content, approximate determinant treatment, saddle selection, fluctuation expansion, and truncation.
5. Take a static or critical route. Statistical Field Theory Preview connects regulated partition functions, order-parameter fields, and effective actions. Combine it with the Phases chapter’s critical exponents, universality, and RG preparation before Critical Phenomena and RG Bridge adds continuum operator matching, fixed points, scaling operators, and relevant deformations.
6. Take the thermal route. Finite-Temperature Methods leads to Finite-Temperature QFT Bridge for Matsubara fields, spectral boundary values, screening, and thermal field-theory continuation.
7. Take the real-time route. Nonequilibrium Dynamics plus source and path-integral preparation leads to Schwinger–Keldysh Bridge for an initial density matrix, doubled contour, in-in observables, influence functionals, and nonequilibrium actions.
8. Take the long-wavelength route. Conserved densities and transport lead to Hydrodynamics and Effective Theory Preview. Schwinger–Keldysh is an advanced continuation, not a universal prerequisite for identifying hydrodynamic variables.
9. Choose further study. Continue on QFT.org suggests further study for each branch.

Example: fermions and an auxiliary spin field

Consider a lattice fermion model with local repulsion and an antiferromagnetic instability. Its lattice-fermion Hamiltonian and a properly regulated coherent-state path integral can be exact representations. A normalized Hubbard–Stratonovich identity can exactly introduce an auxiliary spin field, and choosing an algebraically equivalent decoupling is not itself approximate. Approximation enters when channels are omitted, the determinant is approximated, a saddle is selected, or fluctuations are truncated; exact decompositions can then disagree. A continuum order-parameter theory is a coarse-grained EFT, whether derived by integrating out modes or constructed from symmetry and matching. RG organizes its change with scale.

Continue to a focused field-theory topic

Finite-Temperature Methods supplies imaginary-time and Matsubara calculations. Nonequilibrium Dynamics specifies real-time protocols. Phases, Order, and Criticality supplies the phase and scaling questions that motivate critical field theories.

The field-theory continuation guide connects each of these starting points to further study. You do not need to complete every branch before following one of them.

Common pitfalls

“Second quantization is already relativistic QFT.” It is an operator language that also describes fixed, conserved nonrelativistic particle number. Lorentz symmetry, antiparticles, microcausality, and relativistic vacuum structure require additional input.

“Every object called a field is the same kind of object.” Operator fields, integration variables, auxiliary fields, order parameters, and hydrodynamic variables have different algebras, measures, observables, and validity.

“A path integral makes a model approximate.” A regulated functional representation can be exact; discretization errors, saddle points, truncated actions, and numerical estimates introduce separate approximations.

“A Hubbard–Stratonovich field is the physical order parameter.” The transformation introduces an auxiliary variable. A saddle expectation may represent order only after channel, normalization, symmetry, and fluctuation arguments.

“Euclidean time directly predicts real-time dynamics.” With exact analytic data, continuation is fixed under the stated analyticity conditions. Reconstructing real-frequency spectra from finite or noisy Euclidean data is a separate, ill-conditioned inverse problem.

“Schwinger–Keldysh automatically means an open system.” The closed contour also describes isolated systems. Noise and dissipation arise only after the state, environment, coarse-graining, and influence functional justify them.

“RG discards microscopic detail without a dictionary.” Universality preserves selected long-distance data; operators, metric factors, relevant deformations, and nonuniversal matching coefficients still matter.

“Hydrodynamics follows from conservation alone.” It also needs the complete slow-variable set, local relaxation or closure assumptions, scale separation, and a controlled gradient window.

“A continuum limit removes the regulator automatically.” Couplings and operators require matching or renormalization, and some microscopic information survives in coefficients, anomalies, topology, or irrelevant corrections.

“Emergent Lorentz symmetry can be assumed.” It must be demonstrated in the stated infrared regime—for example, by an infrared fixed point together with the irrelevance or tuning of Lorentz-violating operators—and checked against velocities, operators, and residual anisotropies.

Exercises

Exercise 1: Classify five field objects

Classify (a) a lattice annihilation operator in real space, (b) a fermionic Grassmann variable in a coherent-state integral, (c) a Hubbard–Stratonovich spin field, (d) a coarse-grained magnetization, and (e) a hydrodynamic density. For each, state whether it is an operator or integration variable and where approximation enters.

Solution

The lattice annihilation field is an operator and can exactly rewrite the microscopic model. The Grassmann field is an integration variable whose regulated coherent-state representation can be exact; approximation enters through discretization handling, action truncation, or evaluation. The Hubbard–Stratonovich field is auxiliary and can be introduced by an exact identity, while channel truncation, saddle point, and fluctuation truncation are approximate. Coarse-grained magnetization is an effective collective field after matching and scale separation. Hydrodynamic density represents a conserved slow variable, but its autonomous constitutive dynamics requires closure and a long-wavelength, low-frequency regime.

Exercise 2: Select three continuations

Route (a) a finite-temperature Euclidean correlator whose real-time spectral width is wanted, (b) a driven initial-state problem with in-in observables, and (c) a continuous transition whose critical exponents are wanted. Name the internal leaves and the main inference hazard in each route.

Solution

For (a), use Thermal Green Functions → Spectral Representation → Analytic Continuation; add the Finite-Temperature QFT Bridge only when the correlator belongs to a field theory requiring that continuation. The hazard is treating Euclidean samples as direct real-time data. For (b), use Nonequilibrium Dynamics → Schwinger–Keldysh; the hazard is omitting the initial density matrix, contour/source normalization, or environment assumptions. For (c), use Phases → Statistical Field Theory plus Critical Exponents, Universality, and RG Preview → Critical Phenomena and RG; the hazard is assigning a universality class from symmetry alone without dimension, range, operator content, relevant perturbations, and scaling checks.

References

• A. Altland and B. Simons, Condensed Matter Field Theory, 2nd ed., Cambridge University Press (2010).
• J. Cardy, Scaling and Renormalization in Statistical Physics, Cambridge University Press (1996).
• E. Fradkin, Field Theories of Condensed Matter Physics, 2nd ed., Cambridge University Press (2013).
• A. Kamenev, Field Theory of Non-Equilibrium Systems, Cambridge University Press (2011).
• J. W. Negele and H. Orland, Quantum Many-Particle Systems, CRC Press (1998).
• S. Sachdev, Quantum Phase Transitions, 2nd ed., Cambridge University Press (2011).