Black holes and assumptions

The headline reads New Study Suggests that Black Holes are Growing Faster than Previously Thought.

"The smaller the galaxy, the greater the fraction of stars in these dense, compact clusters," Swinburne researcher Dr Nicholas Scott said. "In the lower mass galaxies the star clusters, which can contain up to millions of stars, really dominate over the black holes."

Previously it was thought that the star clusters contained a constant 0.2 per cent of the galaxy mass.

If you look at the paper in arXiv things look a little different. The first section is devoted to explaining the background for the report--namely that it was well known that there were inconsistencies and different models, and that therefore it was reasonable to revisit the black hole mass to galaxy mass for a particular class of galaxy I'd not heard of before.

It is nice to see science results out there in the news, but one of the things we give up along the way is a sense of proportion. Reporters don't have it for science issues (not for very much else either, if my eyes do not deceive me). For example, how did we wind up with that "0.2 per cent?"

The answer turns out to rely on estimates for the mass of black holes. Astronomers estimated the mass of black holes in different galaxies, noted a pattern, and came up with a model that would explain the pattern. That's the way it is supposed to work; no problems.

Something comes along that deviates from the pattern: Why? If the model is good, maybe something dramatic happened in that galaxy and they can speculate about what (merging galaxies?). Nobody knows better than the astronomers themselves what the hidden assumptions are, and how often the dramatic deviation melts away when the underlying model is tuned up.

But your average--or even above-average--reporter doesn't have the domain knowledge to be able to guess at what's an incipient paradigm shift and what's a wonky assumption in astronomy. I don't, and I'm in an allied field. I kept hearing that there was a deficit of medium-sized black holes, and all I was thinking was "OK, that's odd."

What I should have thought is "How do they measure how big the black holes are; how many medium sizers do they expect; and why do they expect that many?"

That first question is pretty important. How do you estimate a black hole's size, given that we're typically millions of light years away? What other things could be happening that might upset the measurement?

That's what those pesky error bars are there for: the ones that never seem to show up in the news reports. The ones that take as long to calculate as the measurements themselves. The ones that turn an "Oh Wow!" headline into a "Let's Check it Again" one.