Looking at the 60 year history of observations of active galaxies, it is clear that the definition of what they are has strongly influenced the methods of finding them. From our present perspective, many of the techniques used over the past 40 years are not truly appropriate and are more in the line of the famous joke of the drunk looking under the lamp post for his lost car keys. In this chapter, I will use the words Active Galactic Nuclei (alias AGN or quasars) to be the equivalent of radiating supermassive black holes, even though this perspective is very recent.
The difficulty in finding AGN is defining what makes the observed 1 radiation different from that due to other processes, in particular those related to normal stars and stellar evolution (e.g., supernovae). 2 This has often been a process of exclusion; that is, the emission does not resemble that from stars or stellar processes. Dust, high-luminosity emission from starbursts, and the possible effects of unusual types of stars complicate the issue. The strong effects of observing in different spectral ranges also need to be taken into account; for example, at R = 22 there are only 100 "optically-selected" AGN per square degree, but at the equivalent flux level in the 2 - 10 keV X-ray band of 3 × 10-15 ergs cm-2 s-1, there are 1000 deg-2. Finally, it is clear that the "non-stellar" signature has a wide variety of forms that gives rise to the "zoo" of names for active galaxies. The spectral energy distributions, optical emission-line properties (strengths, widths, and nature), line of sight column densities, time variability characteristics, and bolometric luminosities of Seyfert 1 galaxies, Seyfert 2 galaxies, BL Lacertae objects, LINERs (Low-Ionization Nuclear Emission Regions), and quasars (to use the names of the largest samples of objects) are all rather different (see, e.g., Risaliti & Elvis, this volume).
It has taken many years and a large amount of effort to finally come to the realization that all these classes are manifestations of the same underlying physical process - emission from near to a supermassive black hole. However, even today it is not certain if all of these sources are driven solely by accretion, or whether there is also energy extraction from the spin of the black hole (see Armitage, this volume). It is also not clear if the energy production is dominated by radiation, relativistic particle production, or bulk motion of material. It is entirely possible that there is a simple relation between the "names" of the sources and their physical natures, but at present this seems very complex and not unique. As opposed to stellar classifications, there is not a one-to-one relationship between the class of the object and its physical nature. However, there are some clear distinctions: for example, in BL Lacertae objects, the observed radiation is dominated by emission from relativistic particles in a jet in our line of sight, and in Seyfert 2 galaxies, the line of sight to the central source is blocked by large amounts of dust and gas (see Hewett & Foltz 1994 for an earlier review and a detailed discussion of the many systematic effects in quasar surveys).
Before discussing the field in general, it is important to consider what a survey really is. As Hewett & Foltz (1994) point out, there are three types of surveys: (1) those that find objects, (2) those that find objects consistently, and (3) those with well-defined selection criteria that allow probabilities to be assigned for selection as a function of survey parameters. Surveys of the first type are the easiest, since the goal is only to provide sources for study that meet some criteria. Surveys of the second type are homogenous in their properties, but completeness is not important. Many of the issues discussed below are more or less important depending on which type of survey is being performed. It is only surveys of the third type that allow comparisons to be made of different wavelength regimes and different survey techniques. While these surveys are the most scientifically important, they are also the most difficult to do.
1 While very frequently the inferred emitted radiation is rather different from that intrinsically produced by the region around the black hole, the nature of surveys is such that we must rely on the observed properties of these sources in order to find and identify them. Back.
2 To date, all the surveys for AGN have relied on detection of radiation across the electromagnetic spectrum. Perhaps in the distant future we will be able to search for AGN via neutrinos, gravitational waves, or even very high-energy cosmic rays, but this is still quite uncertain. Back.