0
$\begingroup$

In quantum field theories, a local transformation of a scalar field $\phi(x)$ is a transformation that involves the field and its derivatives at same point. See for instance Weinberg's QFT textbok, Vol.1 Sec. 7.7, or arxiv:hep-th/9310042, section 4.1 (it is called point transformation there).

Examples of such transformations are $\phi(x)\to \phi(x)+f(x)$ or $\phi(x)\to \int_y f(x-y)\phi(y)$, with $\hat f(p)$ analytic for small momentum $p$.

Now, under an infinitesimal dilatation $x\to (1+\epsilon)x$, the field transform as $$\phi(x)\to \phi(x)+\epsilon(\Delta_\phi+x^\mu \partial_\mu)\phi(x),$$ where $\Delta_\phi$ is the scaling dimension of the field.

Obviously, the transformation depends explicitly on $x$ in such a way that it is not local as defined above (in momentum space, it involves a derivative wrt to momentum).

My question is: how much "non-local" is the dilatation transformation? Can it be understood as local in a sense?

I ask this because integrated operators nevertheless behave nicely under this transformation, since for instance $$ \int_x \phi(x)^2\to (1+\epsilon(2\Delta_\phi-d))\int_x \phi(x)^2, $$ by integration by part.

$\endgroup$
6
  • $\begingroup$ Where is this definition of local transformation taken from? In most texts in QFT a local transformation is one that the transformation parameters depend on space(-time). E.g. by that definition, a local translation: $\phi(x)\mapsto\phi(x)+c(x)$, for some function $c$ that depends on $x$ is a local transformation. By your definition it wouldn't be because $c$ depends explicitly on $x$. $\endgroup$ Commented May 19, 2021 at 8:03
  • $\begingroup$ you are right, this transformation is of course allowed. I've changed the question to make it clearer. $\endgroup$
    – Adam
    Commented May 19, 2021 at 8:36
  • $\begingroup$ I still don't understand (v3) why $\phi(x)\mapsto \phi(x)+f(x)$ is a point (or local in your definition) transformation. $\phi\mapsto \phi+f[\phi]$ would be, but $f(x)$, no. $\endgroup$ Commented May 19, 2021 at 10:58
  • $\begingroup$ @ɪdɪətstrəʊlə I don't understand... In your first comment, you said that a shift by a function is a local transformation. I agreed and changed the text so that this case is included. And now you ask why it is... Or am I missing something? $\endgroup$
    – Adam
    Commented May 19, 2021 at 12:43
  • $\begingroup$ I said that by the normal, textbook definition of local transformations, it would be. And that your definition of a local transformation is not the same as the textbook one (at least I don't see it). $\endgroup$ Commented May 19, 2021 at 12:48

0

Your Answer

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