It appears that central supermassive black holes are a nearly universal component of galactic bulges. Do the central black holes form first and serve as condensates for galaxies? Or do they build up as galaxies grow and merge? SAFIR will be a powerful tool to answer these questions. The low lying H2 lines at 17 and 28.2um are one of the few ways to study warm molecular gas condensations prior to the formation of metals, for example molecular gas around primordial massive black holes. Line widths and profiles will indicate whether the central mass is highly compact (suggesting a black hole), or if the molecular cloud is just in a mild state of turbulence. At the current epoch, galaxy mergers produce huge far-infrared fluxes through a combination of violent starbursts and AGNs associated with their central black holes. The distinction of starbursts from supermassive black holes as power sources for young and distant galaxies was identified in the Decade Report (NAS Decade Report 2001) as a major goal for new far-infrared telescopes. What happens during the much more common mergers that build galaxies in the early Universe? COBE showed that the far-infrared-to-submillimeter energy density in the early Universe is comparable to that in the visible-to-near-infrared. What are the relative roles of dust-embedded AGNs and starbursts in producing this luminosity? Do AGNs at high redshift differ in basic properties from nearby ones? Models of the cosmic X-ray background indicate that the great majority of AGNs at high redshift are heavily absorbed (Gilli, Salvati, & Hasinger 2001; Comastri et al. 2001). Thus, these answers must be sought in the far-infrared where optical depths are low (interstellar medium, ISM, optical depths are similar at 20m and 6 kev and rapidly decrease at longer infrared wavelengths and higher X-ray energies). The fine structure lines of NeII (12.8um), NeIII (15.6um) and NeV (14.3um) are the best tool to distinguish unambiguously whether the ISM of a dusty galaxy is ionized by a starburst or by an AGN. Not only are the line ratios very well separated, but their extinction is reduced by more than a factor of thirty compared with the visible. At the epoch of peak quasar activity, these lines will be redshifted to the 45 to 55m range. Figure 1 shows the spectrum of the nearby active galaxy, Centaurus A, illustrating the wealth of spectroscopic features that will be redshifted into the 25 - 100um spectral region in early galaxies. Using these fine structure lines, a 10m far-infrared telescope will have the necessary resolution and sensitivity to determine the roles of star formation and nuclear activity in the early Universe. The full suite of infrared fine structure lines probes a very wide range of excitation energy, allowing SAFIR to constrain the UV spectra of AGNs and extending work done with ISO on a few nearby Seyfert galaxies to large lookback times. In addition, many of these lines have relatively high critical densities (up to ~10^10 cm^-3), so they have a unique ability to probe the density of the gas around AGNs. The angular resolution of SAFIR is a critical contribution to these studies, and complements that of forthcoming NASA investments in telescopes operating at shorter wavelengths. This capability is also illustrated in Figure 1 with a comparison of the SAFIR and SIRTF beams on a simulated JWST image.