0
$\begingroup$

I am trying to understand the textbook derivation of the two-point function in CFT as spelled out in this CFT book by Francesco. The same derivation can also be found in this thread.

However, I don't understand the final step they do in their approach.

So the derivation goes something like this:

Consider two primary fields $\phi_1(x_1)$ and $\phi_2(x_2)$. Since we know how the primary fields transform we can write their two-point function as

\begin{equation} \langle\phi_1(x_1)\phi_2(x_2)\rangle =\left|\frac{\partial x'}{\partial x}\right|^{\Delta_1/d}_{x=x_1} \left|\frac{\partial x'}{\partial x}\right|^{\Delta_2/d}_{x=x_2} \langle\phi_1(x_1')\phi_2(x_2')\rangle. \end{equation}

Now, from translational and rotational symmetry we have that $$\langle\phi_1(x_1)\phi_2(x_2)\rangle = f(\left| x_1-x_2\right|).$$

By scaling $x$ i.e. letting $x\rightarrow \lambda x$ we get that

\begin{equation} f(\left| x_1-x_2\right|) = \lambda^{\Delta_1 +\Delta_2}f(\lambda \left| x_1-x_2\right|). \end{equation}

Now comes the "final" step that I fail to see or understand. Without too much explonation (maybe because it is straight forward, but I still don't se it) they say that this fixes the two point function up to an overall constant:

\begin{equation} \langle\phi_1(x_1)\phi_2(x_2)\rangle = \frac{C}{|x_1 - x_2|^{\Delta_1 + \Delta_2}}. \end{equation}

I don't see how letting

\begin{equation} \langle\phi_1(x_1)\phi_2(x_2)\rangle = f(\left| x_1-x_2\right|) = \lambda^{\Delta_1 +\Delta_2}f(\lambda \left| x_1-x_2\right|) \end{equation} tells us that we must have $f(x) \approx \frac{1}{x}$.

If I use this ansatz I get

\begin{equation} f(\left| x_1-x_2\right|) = \frac{1}{\left| x_1-x_2\right|}= \lambda^{\Delta_1 +\Delta_2} \frac{1}{\lambda \left| x_1-x_2\right|} = \lambda^{\Delta_1 +\Delta_2 -1} \frac{1}{ \left| x_1-x_2\right|}. \end{equation}

The only way I could see that this is true is if it was known that the sum of the scaling dimensions must always be one i.e. $\Delta_1 + \Delta_2 =1$, but I don't know if this is true?

Anyone who see my confusion?

$\endgroup$
1
  • $\begingroup$ Shouldn't the scaling property suggest that $f(x)\sim 1/x^{\Delta_1+\Delta_2}$? $\endgroup$
    – Meng Cheng
    Commented Jan 26, 2022 at 12:38

1 Answer 1

1
$\begingroup$

As $f( |x|)= \lambda^\Delta f(\lambda |x|)$ holds for any $\lambda$, take the particular value $\lambda= |x|^{-1}$. Then the relation becomes $$ f(|x|)= |x|^{-\Delta}f(1) $$ Call $f(1)$, "$C$".

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.