Roeland van der Marel of the Institute for Advanced Study in Princeton, N.J., has been investigating this question. He responds:
Quasars were discovered in the 1960s. The name is a contraction of 'quasi-stars' or 'quasi-stellar object,' which refers to the fact that quasars are generally point sources (like stars) that have no observable extent. The radiation from quasars is greatly stretched, or redshifted, which implies that these objects are very far away--the stretching of their light is presumed to result from the overall expansion of the universe. To be visible to us despite their tremendous distances, quasars must produce a tremendous amount of energy, yet they must be small enough to appear pointlike.
Some galaxies contain very bright, compact regions in their centers; these objects are known as active galaxies. Quasars have spectral properties that are very similar to those of the central regions of these active galaxies. Astronomers have therefore long believed that quasars are galaxies in which only the bright nucleus can be seen. Recent observations with the Hubble Space Telescope have confirmed this hypothesis: high-contrast images clearly show, for the first time, the galaxies within which quasars are embedded.
The most plausible mechanism that could produce the vast energy output of both quasars and active galaxies is the accretion of matter onto a massive black hole. As gas falls into the black hole, it forms a disk that grows extremely hot because of friction; this process releases energy in accordance with Einstein's equation E = mc2. This standard model of how quasars work is supported by very convincing evidence indicating the presence of massive dark objects in the nuclei of some active galaxies. As a result, a substantial majority of professional astronomers is convinced that the violent behavior of quasars is indeed caused by black holes.
Now onto the second part of the question, 'Does the discovery of black holes in the centers of many normal galaxies mean that all galaxies were once quasars?' Nearly all quasars are seen at high redshifts. Hence, we see them when they were still very young: they are very distant, so their light took a very long time to reach us. In the current epoch, quasars are extremely rare. Quasars therefore must evolve into something else that we do see in large abundance today (that is, in the relatively nearby portion of the universe). The most likely objects that they evolved into are normal, quiescent galaxies. If this is the case, then normal galaxies should still have massive black holes in their nuclei, like quasars, but these black holes should now be starved of fuel.
Observations of several normal galaxies, including our own Milky Way,
have provided evidence for the presence of such black holes. What we
do not yet know is what fraction of all normal galaxies once passed
through a quasar phase. To answer that question, we need high-spatial
resolution spectroscopic observations of a large sample of 'normal'
galaxies, which would reveal the presence of massive central black
holes. This kind of information can now be obtained using the STIS
spectrograph, which was recently installed by NASA astronauts on the
Hubble Space Telescope. Several groups will try to answer the question
in the coming years.