On this page:
1 Introducing additional fields
2 BRST operator is nilpotent off-shell
3 Form
3.1 For standard family of Lagrangian submanifolds
3.2 Generic family of Lagrangian submanifolds
3.3 Equivariant


Bosonic string quartet

We will now show how to integrate over metrics keeping the BRST operator nilpotent off-shell

    1 Introducing additional fields

    2 BRST operator is nilpotent off-shell

    3 Form

      3.1 For standard family of Lagrangian submanifolds

      3.2 Generic family of Lagrangian submanifolds

      3.3 Equivariant

1 Introducing additional fields

Let us return to our solution of Master Equation given by Eq. (1). Let us add the “BRST quartet” and modify Eq. (1) by adding :

(13)

The standard choice of Lagrangian submanifold:

leads to degenerate action funcional. Now let us shift it using the gauge fermion:

(14)

This gives the family of Lagrangian submanifolds:

2 BRST operator is nilpotent off-shell

It is straightforward to construct the new Darboux coordinates compatible with the new Lagrangian submanifold; all we need to do is to shift:

In terms of these new variables:

(15)

The first line does not depend on antifields, and the second line is linear in antifields; we omitted the “” labels on antifields.

We observe that:
  • is the Lagrange multiplier for

  • plays the role of -ghost

  • The Master Action of Eq. (15) depends on antifields only up to first power

Therefore, the BRST operator is nilpotent:

3 Form

Since is nilpotent off-shell, we can use our procedure for constructing in BRST formalism.

3.1 For standard family of Lagrangian submanifolds

For the standard family, generated by the family of gauge fermions parametrized by as in Eq. (14), our formula gives:

(16)

where is the first line of Eq. (15).

3.2 Generic family of Lagrangian submanifolds

For the general family we have to use our general formula; there is no further simplification.

3.3 Equivariant

We will define diffeomorphisms in the usual way:

For the case of standard Lagrangian submanifold, of Eq. (16) is already base. For the general case, we have to use our general formula for equivariant in BRST formalism with . There is nothing particularly specific to bosonic string, except for:

Notice that there are two kinds of symmetries: diffeomorphisms parametrized by and -shifts parametrized by arbitrary -independent functionals .

That should be -independent follows from

It seems that we now have a larger symmetry group: instead of just diffeomorphisms we have diffeomorphisms plus -shifts. This is probably related to adding additional fields.