Galaxy birth and development is a never-ending process, and clues to early black hole evolution are spread throughout our own galaxy and around the universe. Astronomers therefore examine modern-day cosmic creatures for clues to their ancestral roots.

If the mightiest black holes indeed developed out of the garden variety, then there ought to be some evidence lying around our cosmic backyard in the form of middleweight versions, one line of thinking goes.

"The existence of middleweight black holes is one of the big unanswered questions in this field," said Cambridge's Haehnelt. "The recent claimed detections are still very controversial."

Regardless, most experts agree middleweights would represent, at best, pocket change to the fully grown black hole, something like Microsoft's initial millions in annual revenue compared to the billions that poured into its coffers during the tech boom.

Researchers on both sides of the middleweight argument mostly agree that the bulk of a jumbo black hole doesn't come through early mergers. Once a critical mass is achieved -- and this appears to coincide with a point in time prior to what astronomers can see today -- a black hole seems to gain most of its mass by swallowing gas from its environment.

Amid all the squabbling over middleweights looms the likelihood of much larger merger candidates.

Galactic mergers take millions of years, so they can't readily be observed in progress.

Other indications of mega-mergers come from relatively nearby quasars.

Richard Larson, a Yale astronomer who studies star formation in galactic nuclei, says galaxies can go through several quasar phases during their lives. In studying quasars at more reasonable distances (which also means not so far in the past), he consistently sees signs of recent galaxy mergers or other large-scale interactions that served as triggers.

"Interactions and mergers are an excellent way to dump a lot of gas into the center of a galaxy," Larson explains. "The first thing this gas does is suddenly form huge numbers of stars."

Bursts of intense star formation seem to last about 10 million to 20 million years around a typical quasar.

Some of the gas that does not go into generating stars falls on in to the black hole. This violent phase of consumption is the one that is readily observed, because the castoff energy turns the incoming gas and dust into a glowing cloud. Eventually, the chaos settles and the new stars become visible. Later, the quasar itself is left naked. Finally, it goes dormant.

Larson figures this scenario for black hole feeding probably applies to the most distant quasars, too. And it supports the notion that black holes do in fact gain most of their bulk by accreting gas.

To sort out the specifics of co-evolution, astronomers will need to see more of the universe and inspect it in greater detail. The prospects are good, especially toward the end of this decade.

A project called LISA (Laser Interferometer Space Antenna) would search for "gravitational waves" kicked up in the aftermath of black hole mergers, perhaps proving that such colossal collisions do occur. The NASA satellite is tentatively slated for launch in 2008.

A vastly improved understanding of dark matter is also needed. Several telescopes should contribute to this effort, but since no one knows what the stuff is, forecasting any sort of resolution is highly speculative.

And the specific mechanics of black holes must be investigated fully. For now, theorists don't even know exactly how matter is shuttled inward and consumed. Much of this work can be done by observing the nearby universe.

Roger Blandford, the Caltech theoretician, has suggested a novel way to prove that early mergers were not serious contributors to black hole growth. Blandford says two primary parameters characterize black holes. Mass is the most obvious. A more subtle measurement is spin.

But if spin can be proved a universal aspect of black holes, then the rate of spin can be used to infer something very important about a black hole's history.

"If black holes grow by merging, by combinations of black holes, they should spin down quite quickly," Blandford explains. "This then becomes a fairly good argument that, if you can show that black holes really are spinning rapidly, they probably didn't grow by merging, but would have grown by accreting gas."

Most important, vision simply must be extended further back in time, beyond the quasars that are now being studied, says Karl Gebhardt, a University of Texas astronomer and a member of Richstone's team.

"They're essentially the tip of the iceberg," Gebhardt says of the objects so far observed. "We are projecting from what we see in a very special number of objects to the whole sample. That is part of the problem of the uncertainty now."

It is ironic to think that when JWST goes up, many astronomers and cosmologists will be banking on black holes to light the way to a scientific account of the earliest epoch of the visible universe, an obscure time they have long dreamed about and can now, almost, see.