"Power exercised in secret, especially under the cloak of national security, is doubly dangerous" US Senator, W. Proxmire
Because the luminous collimated radio emission must have been generated in their cores, the nuclei of all HzRGs must have been active at some time in their histories.
5.1. Hidden or dormant quasar - The Unification Church
The most extreme form of nuclear activity is seen in quasars, for which the nuclei are so bright that the galaxies are observed as "quasi-stellar objects". There is strong evidence that radio galaxies and quasars are manifestations of the same parent population. When we observe objects to be radio galaxies, the associated quasars are believed to be either (i) hidden due to our special viewing angle and/or (ii) highly variable and observed when they are in a dormant state.
The most widely accepted models for unifying quasars and radio galaxies postulate that their observed properties are dependent on the orientation of the radio source axes relative to us, the observers. For a review of unification via orientation see Antonucci (1993), Urry and Padovani (1995), Urry (2004). One of the key factors in establishing the validity of HzRG unification through orientation has been the signature of a hidden quasar detected in some radio galaxies in polarized light. Polarimetry with the Keck Telescope (Vernet et al. 2001) (See Section 3.5.2) show that HzRGs are usually, but not always, highly polarised just redward of Ly and that dust-reflected quasar light appears generally to dominate their rest-frame ultraviolet continua. Furthermore, broad emission lines, with equivalent widths similar to those seen in quasars, are detected in the two most polarized HzRGs in the Vernet et al. (2001) sample of nine objects.
There is a convincing case that an obscured quasar exists in those HzRGs for which high fractional polarisations and broad polarised recombination lines have been measured. What about HzRGs with low polarisation? For these objects, it is possible that active quasars are present that are even more highly obscured than the polarised HzRGs. However an equally plausible explanation is that the quasars are dormant during the epoch of observation (van Groningen et al. 1980). We know that quasars are highly variable, varying by as much as 2 magnitudes during a day. Quasar variability has been studied on timescales of up to 50-years (de Vries et al. 2006). The variability appears to be intrinsic to the AGNs, and occurs in flares of varying time-scales, possibly due to accretion-disk instabilities.
During the lifetime of an extended radio sources (at least a few × 107 y), quasar activity can be expected to vary considerably. Knots in the radio morphologies can be interpreted as tracing bursts of quasar activity in the radio nuclei (Miley 1980). According to such a dormant quasar model, if the flare duration is substantially shorter than the quiescent periods the objects would be more likely to be observed as radio galaxies than radio-loud quasars. Also if the flares recurred on a timescale shorter than the recombination time (~ 104 - 105 y), the excitation of the gas would persist. A recurrently flaring quasar could also be a source of excitation for some of the non-radio Ly halos (Section 3.2.5).
In summary, there is strong evidence that quasars exist in the nucleus of all HzRGs. Although orientation and varying intrinsic activity both probably contribute to the observed properties, their relative importance is still unclear.
5.2. Supermassive black hole - Powerhouse of the AGN
Extragalactic radio sources, and all active galactic nuclei, are believed to be powered by gravitational energy produced by rotating supermassive black holes (SMBHs) in their nuclei (e.g. Lynden-Bell 1969). Material in the galaxies is accreted onto the SMBHs, converted into kinetic energy and ejected as collimated relativistic jets along the rotation axes of the SMBHs to produce the radio sources. For a theoretical review of this field, see (Blandford 2001). Although there is wide agreement about the general idea, many of the details are not well understood.
Observational evidence for the widespread occurance of SMBHs in galactic nuclei include (i) spectroscopy of the gigamaser in the nucleus of NGC 4258 (Miyoshi et al. 1995) (ii) spectroscopy and photometry of galactic nuclei with the HST leading to the correlation between the masses of galaxy bulges and SMBHs in early type galaxies (e.g. Magorrian et al. 1998, Kormendy and Gebhardt 2001, Häring and Rix 2004, Novak et al. 2006), (iii) the small collimation scale of radio sources, measured with VLBI, (1017 cm (Junor et al. 1999)) and (iv) the similarity of the large-scale and small-scale orientation of some giant radio sources, indicating that the "memory" of collimation axis can persist for as long as ~ 108 y (Fomalont and Miley 1975, Schilizzi et al. 1979, Miley 1980).
Because of the interaction of the radio jet with the gas, line widths in the central regions of HzRGs cannot be used to derive the masses of the presumed SMBHs in their nuclei, as for radio-quiet galaxies. The empirical correlation of SMBH with the bulges of early type galaxies (Häring and Rix 2004) would predict black hole masses of ~ 109 M at the centre of a 1012 M HzRG. (McLure and Jarvis 2002) have derived similar masses for SMBHs in radio-loud quasars from the width of their emission lines. However, such estimates should be treated with caution. There is a possibility that the widths of the emission lines are widened due to pressure from the radio jets Also, little is known about how the SMBHs assemble during the formation and evolution of a massive galaxy, such as a HzRG, so extrapolation of results from lower redshifts may not be valid. Silk and Rees (1998) have postulated that supermassive black holes form within the first sub-galactic structures that virialise at high redshift, and are in place before most galactic stars have formed.
5.3. Extinction of the "dinosaurs"
Luminous quasars and radio galaxies have undergone strong evolution in their space densities since z ~ 2 (the peak of the "quasar era") and both species are now virtually extinct (Section 1.4). The reason why such extreme manifestations of nuclear activity are no longer with us is not fully understood. One obvious explanation is that the supply of "food" for the SMBHs began to diminish as more and more of the the gas in massive galaxies was converted to stars (e.g. Menci 2004).