Dubious Deductions from AGN Survey Data

The participants in this meeting are almost all carrying out the hard work of making many different types of AGN surveys. Since it’s so much easier to criticize other people’s work than to do actual work myself, I’ll just present some demurs regarding recent papers drawing conclusions from various AGN survey data. In particular I’ll mention some questionable interpretations of surveys of Seyfert 2 near-UV polarization; interpreting the results of searches for polarized broad H-alpha lines in Seyfert 2s; testing the beam model for radio galaxies and quasars; testing the unification of Seyfert spectral types with a torus; and finally testing the energy sources for Ulirgs, especially those with Liner optical spectra. Only the polarized broad H-α results are examined in detail here. 1. Levels of polarization of the nuclear light in Seyfert 2 galaxies Koski (1978) showed that Seyfert 2s and Narrow Line Radio Galaxies generally have UV excesses relative to the SEDs for the Pop II stars which dominate the optical light in arcsec apertures. He showed the spectra could be parameterized approximately by a normal Pop II SED, plus a suitably normalized power law which dominates in the near-UV. The brightest powerful Seyfert 2 nucleus, that of NGC1068, can indeed be fit quite well that way (Miller and Antonucci 1983, McLean et al 1983, Antonucci et al. 1994), and furthermore the power-law component has a wavelength-independent polarization of ∼ 16% in arcsec apertures. It is crucial to note that the broad permitted lines show the same percent P as the power-law continuum component accounting for the “UV excess.” This strongly suggests that the scattering paths of the broad lines and the power law are very similar geometrically. Finally, the permitted lines have normal equivalent width in polarized flux. This confirms that the hidden source is simply a normal Seyfert 1 nucleus, whose light is scattered into the line of sight. The polarization angles of NGC1068 and other Seyfert 2s lie perpendicular to the radio axes, so that the photons’ last flights before scattering into the line of sight were along the axes. Thus the equatorial directions must be opaque, and the torus model was born. In principle the polar scattering could result from intrinsically anisotropic emission, but the anisotropy would need to be The Proceedings of the Colloquium will be published by the Astronomical Society of the Pacific as a volume of the ASP Conference Series. The editors will be Richard Green (Editor-in-Chief, USA), Edward Khachikian (Armenia) and David Sanders (USA).