In Blumenhagen's book "Introduction to Conformal Field Theory", I found the statement

The algebra of infinitesimal conformal transformations in an Euclidean 2-dimensional space is infinite dimensional.

He concludes this after finding that the generators of the infinitesimal conformal transformations (i.e. a basis for its Lie algebra) are $l_n=-z^{n+1}\partial$ and $\bar l_n= -\bar z^{n+1}\bar\partial$ for $n\in\mathbb N$.
My problem with this is: shouldn't the elements of the Lie algebra be linear combinations (with real coefficients) of the basis elements $\partial,\bar\partial$ of $T_eG$? (Instead of a roduct of the derivatives with polynomials). Furthermore (and related to the previous point), if our Lie group is 2-dimensional (as a manifold), that implies that its Lie algebra is 2-dimensinal (as vector space), so definitely not infinitely dimensional. What is going on here?


You are confusing 2 different things:

A d-dimensional Lie group has a d-dimensional algebra which is the tangent space at the identity.

Here, there is an infinite dimensional group acting on a 2 dimensional space. There is a representation of the algebra as vector fields (derivations) on this space, given by the formulas you stated. (infinitesimal conformal transformations)

The generators should depend on the point z, as each conformal transformations act differently around each point - for example, the map $z$ to $z^2$ acts infinitisimaly as $z+\delta z$ to $z+2z\delta z$. The dependence on the point is easiliy understood - for example under this map, the more far z is away from the origin, it is stretched more.

  • $\begingroup$ I suppose the infinite dimensional group acting on the 2-dimensional space is the "Infinitesimal conformal group"? (And not the Conformal group itself). This is vague for me since so far I only understood what Blumenhagen did with infinitesimal conformal transformations as transformations of the conformal group near the identity $\endgroup$ – Soap May 5 '18 at 19:49
  • 1
    $\begingroup$ No. It is the conformal group acting on the 2-d space. Upon exponentiation, the generators will become elements of the conformal group itself. The generators themself are not acting on the space, only their exponents do. $\endgroup$ – tsufli May 5 '18 at 19:53
  • $\begingroup$ Ok, but how do "infinitesimal conformal transformations" come into play? From what I understood so far, we have: the conformal group $G$ together with a representation of $G$ in a 2D euclidean space. $G$ has a Lie algebra which has a basis given by the generators that I wrote above. What now? $\endgroup$ – Soap May 5 '18 at 20:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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