On Monday we leave for a 2-day meeting in Brussels about polarization in AGN. This is a good place to understand (if anybody does) how does polarization occur in AGN. The present paradigm of Seyfert galaxies is that Type-1 and Type-2 are exactly the same object. Type-1 Seyferts are characterized for the observation of broad lines (expected to occur very close to the black hole that serves as an engine of the AGN) together with narrow lines occurring well above the black hold. Both kinds of lines are observed simultaneously on the intensity spectrum. Type-2 Seyferts are characterized for the observation only of narrow lines. Broad lines are not conspicuous in these objects. 

The present unification model identifies Type-1 sources when the AGN is observed face-on, while Type-2 happen when the AGN is observed edge-on, as shown in the following figure.

Although the broad line region (BLR) is not observed in type-2 sources, they are observed in polarized light, as a consequence of the scattering in the electron clouds above the black hole. Given that scattering in electrons does not depend on wavelength, the clouds behave essentially as a mirror, in which the "reflected" light is polarized.

The central engine is expected to be surrounded by a dusty torus. Initially, it was assumed to be a smooth distribution of dust, but a clumpy dusty torus gives better results in terms of understanding the observations. If this dusty torus is clumpy, there is a point we still do not understand. In principle, it is possible that the light from the central engine also suffers scattering in the dust clumps and induces polarization, which would be observed preferentially in the mid-infrared. Does it make sense to calculate/observe the spectropolarimetric characteristics of the scattered radiation? We'll see if we can answer this question in the meeting.