research35 min studyLesson 735 in Physics

Conformal blocks and Casimir equations

Package descendants into blocks while fixing the channel, eigenvalue and small-distance branch.

On this page

What you’ll understand

  • Identify the family summed by a global conformal block.
  • Compute quadratic Casimir eigenvalues with a declared convention.
  • Check one exact chiral block against its convergent series.

A conformal block is the contribution of one exchanged primary and all its conformal descendants to a chosen correlator channel. It separates representation kinematics from the CFT's dimensions and OPE coefficients. The same dimension and spin give the same block only after the spacetime dimension, external dimensions, tensor structures and normalization convention are fixed. A finite block calculation is therefore not a computation of which primaries actually occur in an interacting theory.

Fix the correlator and its normalization

Use four identical real scalar primaries in Euclidean dimension d at least three, with unit two-point coefficient and the reduced four-point function defined by the preceding distance prefactor. In the channel that fuses points one and two, each exchanged symmetric traceless primary contributes its normalized block. For identical bosonic scalars, only even integer spin appears in this OPE. Different global-symmetry sectors or nonidentical external operators require additional coefficients and channel bookkeeping.

Choose the identity block to be one and scalar exchanged blocks to have leading behavior equal to u to half their dimension in the simultaneous OPE limit. Spinning blocks need a fixed leading angular polynomial and tensor normalization as well; a positive rescaling can be absorbed consistently into the coefficient convention. Mixing a block formula from one convention with coefficients from another changes the correlator. The full family, rather than an isolated leading power, is what transforms consistently under channel comparisons.

G(u,v)=∑OλϕϕO2gΔ,ℓ(u,v),g0,0=1,gΔ,0∼uΔ/2(u→0, v→1)\mathcal G(u,v)=\sum_{\mathcal O}\lambda_{\phi\phi\mathcal O}^2g_{\Delta,\ell}(u,v),\qquad g_{0,0}=1,\qquad g_{\Delta,0}\sim u^{\Delta/2}\quad(u\to0,\ v\to1)
The displayed positive squared-coefficient form assumes normalized real operators and reflection positivity. Spin-dependent leading normalization is fixed consistently throughout a chosen block basis.↔ Scroll wide equations horizontally.

Use the Casimir equation with boundary data

Apply the quadratic conformal Casimir built from generators acting on the first pair of insertions. Every state in an irreducible exchanged family has the same Casimir eigenvalue, so its block is an eigenfunction of a differential operator in cross-ratios. This turns a sum over descendants into a differential-equation problem. The eigenvalue depends on the exchanged dimension and rotation representation, while the differential operator also knows the external dimensions and spacetime dimension.

The eigenvalue equation alone does not select a unique physical block. One must impose the OPE asymptotic behavior and the proper branch, distinguishing the desired dimension from its shadow partner. Singular solutions or wrong boundary conditions can share the same eigenvalue. The Casimir is not a positive Hamiltonian or a state norm, so a negative eigenvalue is not itself a unitarity violation. Positivity constraints must instead be applied to the radial representation and its descendants.

D12gΔ,ℓ=CΔ,ℓgΔ,ℓ,CΔ,ℓ=Δ(Δ−d)+ℓ(ℓ+d−2)\mathcal D_{12}g_{\Delta,\ell}=C_{\Delta,\ell}g_{\Delta,\ell},\qquad C_{\Delta,\ell}=\Delta(\Delta-d)+\ell(\ell+d-2)
The convention defines the Casimir differential operator with this eigenvalue, without an extra factor of one half. Δ, ℓ and the eigenvalue are dimensionless; ℓ labels a symmetric traceless rotation tensor.↔ Scroll wide equations horizontally.

Benchmark an exactly summable simpler block

In two dimensions the global conformal algebra factorizes into holomorphic and antiholomorphic parts. A holomorphic global block with identical external weights can be written with a hypergeometric function. This is a useful analytic benchmark, but a global block includes only global descendants and generally differs from a full Virasoro block. The central charge and the extra stress-tensor descendants enter the latter and cannot be inferred from a global expression alone.

At a particular exchanged holomorphic weight the hypergeometric series reduces to a logarithm, allowing a direct truncation check. Its branch is fixed initially in the real interval between zero and one; continuing past singularities requires a prescription. The function lab can show a simple polynomial and its derivative, but it does not solve Casimir equations, choose a shadow branch or generate CFT blocks. Any polynomial used as a local approximation needs an error estimate appropriate to the actual block.

k2h(z)=zh 2F1(h,h;2h;z),k2(z)=−ln⁡(1−z)=∑n=1∞znn(∣z∣<1)k_{2h}(z)=z^h\,{}_2F_1(h,h;2h;z),\qquad k_2(z)=-\ln(1-z)=\sum_{n=1}^{\infty}\frac{z^n}{n}\quad(|z|<1)
This is a two-dimensional holomorphic global block in the stated identical-external-weight normalization. h = 1 gives k2; it is not the complete two-dimensional four-point function or a general Virasoro block.↔ Scroll wide equations horizontally.

A Casimir is not a positive energy

In d = 3, compute the chosen Casimir eigenvalue for an exchanged symmetric traceless primary with Δ=4\Delta=4, ℓ=2\ell=2, and for a scalar with Δ=1\Delta=1. Does the second value alone violate positivity?

  1. For the spin-two exchange, the dimension contribution is 4(4−3)=44(4-3)=4 and the spin contribution is 2(2+3−2)=62(2+3-2)=6. The total is ten.

  2. For the scalar, spin contributes zero and the dimension term is 1(1−3)=−21(1-3)=-2. A Casimir eigenvalue can be negative even though a physical state has positive norm.

  3. The scalar dimension one is above the d = 3 nonidentity scalar unitarity bound one half. The negative Casimir does not exclude it; actual existence and its OPE coefficient remain dynamical questions beyond this representation check.

The result

The eigenvalues are 10 and −2; the negative scalar value is not by itself a positivity violation.

Explore the model

Interactive lab

Interactive lab · function

A polynomial baseline is not a block solver

Reset and set BOTH Coefficient a = 1 and Position x = 0.50; predict y = 0.25 and slope 1. Reset and set a = 1 and x = 1; predict y = 1 and slope 2. This exact quadratic supplies a simple derivative comparison, not a Casimir eigenfunction, conformal block or truncation certificate.

A polynomial baseline is not a block solverThis lab uses only y = ax². Playback traces the curve within the fixed integration interval from 0 to x; it does not estimate numerical error or change the chosen point. Current readings: y at chosen x: 0.608; Tangent slope · 2ax: 1.35; Signed area · ax³/3: 0.182. All parameters are adjustable with the labeled controls below.−2.4−1.201.22.4−9−4.504.59xy(x, ax²)
y = ax²TangentSigned integral
y at chosen x
0.608
Tangent slope · 2ax
1.35
Signed area · ax³/3
0.182
t = 0 s

This lab uses only y = ax². Playback traces the curve within the fixed integration interval from 0 to x; it does not estimate numerical error or change the chosen point. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.

Look forSolving a differential equation requires its operator, normalization and boundary data.

Check your understanding

What additional information is needed besides a Casimir eigenvalue to select the intended conformal block?

Try it yourself

Evaluate the holomorphic global block k2(1/2)k_2(1/2) exactly. Compare it with the first two nonzero series terms and bound the omitted positive tail using 1/n≤1/31/n\le1/3 for n at least three.

Take a moment to make a prediction or work it out on paper. Open a hint when you need one.

Give me a hint

The first terms are z and z²/2. Sum a geometric upper bound for all powers beginning with z³.

Compare with the solution

The exact value is ln⁡2≈0.693147\ln2\approx0.693147. The first two terms give 1/2+(1/2)2/2=0.6251/2+(1/2)^2/2=0.625, so the actual remainder is about 0.0681470.068147. A positive upper bound is ∑n=3∞(1/2)n/3=(1/8)/[3(1−1/2)]=1/12≈0.083333\sum_{n=3}^{\infty}(1/2)^n/3=(1/8)/[3(1-1/2)]=1/12\approx0.083333. The exact remainder lies below it. This validates this particular chiral global-block series at the chosen point, not an arbitrary Virasoro block or a bootstrap spectral truncation.

Take it with you

  • A block is one full exchanged conformal family in a fixed channel.
  • Casimir eigenvalues require OPE boundary data to select solutions.
  • Global and Virasoro blocks sum different descendant algebras.

Read further

Original sources and deeper explanations to follow at your own pace.

Optional progress, saved in this browser when available. Revisit any lesson at any time.

Lesson not marked complete.