Cosmic Star Formation: When Elegant Theories Are Wrong
An astronomer wrote about “cosmic train wrecks” in Science recently.1 Paolo Coppi (Yale) was speaking about galactic mergers, but he could have just as well been talking about current cosmological models. Things once thought to be understood are coming in for new scrutiny, now that more powerful telescopes can peer deeper into the veiled hearts of galaxies. One galaxy in particular, NGC 6240, thought to be the result of a merger, was mapped recently in unprecedented detail.
In the middle of a rather straightforward article describing current thinking about what happens when galaxies collide, how stars form, and how black holes behave, he ended one paragraph with a surprise. It was kind of like the ending word “not” in the slang of young people – e.g., “Astronomers understand star formation – NOT!”
Detailed observations of nearby galaxies, the only kind we could carry out until recently, identified two main modes of star formation: powerful and rapid “starbursts” caused by NGC 6240-like collisions and the much less dramatic but quasi-steady formation seen in the disk of our Galaxy. Because objects like NGC 6240 are rare today, one might speculate that most stars form “quietly” in disks. The larger, so-called elliptical galaxies, which do not contain much gas, then come from late-time mergers of smaller disk-dominated galaxies that have turned their gas into stars. Mergers play a minor role, mainly gravitationally scrambling already-made stars. While elegant, this story seems wrong.
The problem is that now it appears most star formation appeared early in the history of the universe. NGC 6240, with two black holes apparently orbiting its center, and no star formation going on today, may be a “common oddball,” – something that should have been rare, but appears to be representative of the state of the early universe. Coppi called this “very surprising” and something that creates an “intriguing new problem for us” –
Today’s elliptical galaxies are “red and dead” because they contain predominantly old (red) stars and are not forming new ones. Very surprisingly, some of the elliptical progenitors also appear to be “red and dead”. Unless we invoke a new mechanism that rapidly and permanently stops star formation, the most massive objects in simulations turn out to be too massive and never sufficiently red and dead.
One solution is to include feedback from the accretion of a supermassive black hole in the models. There seems to be observational support for actively-accreting black holes in systems like NGC 6240, with regions of active star formation going on. “This plus the surprising discovery that every nearby elliptical galaxy contains a black hole with a mass proportional to that of the galaxy strongly hints that rapid star formation and rapid black-hole feeding and growth are both inevitable and closely connected consequences of a cosmic train wreck like NGC 6240 where gas is gravitationally squeezed into a very small volume.” But where does the language of observation get distinguished from theory in such a statement?
From that point on, Coppi focused on prospects for improved observations. The Laser Interferometry Space Antenna (LISA), expected to be operational in 2015, might be able to detect the signature of black hole mergers through gravitational waves they emit. But there is “considerable speculation,” he said, about whether black holes accrete slowly by feeding on their own stars, or form catastrophically through mergers of galaxies. He’s not even sure LISA would be able to tell.
In his discussion, Coppi was assuming black holes are real. Better not tell him about other astronomers who are denying that black holes even exist. A recent article in ScienceNOW Daily News began,
If new calculations are correct, the universe just got even stranger. Scientists at Case Western Reserve University in Cleveland, Ohio, have constructed mathematical formulas that conclude black holes cannot exist. The findings–if correct–could revolutionize astrophysics and resolve a paradox that has perplexed physicists for 4 decades.
There’s no doubt that very massive, compact objects exist in the centers of many galaxies. Asked what to do with these observations, which lead most astronomers to believe the universe is full of black holes, “‘[Lawrence] Krauss replies, ‘How do you know they’re black holes?” No one has actually seen a black hole, he says, and anything with a tremendous amount of gravity–such as the supermassive remnants of stars–could exert effects similar to those researchers have blamed on black holes.”
Krauss and colleagues performed detailed calculations taking into account the relativity of time. They showed that time stops before a singularity forms, meaning “black holes can’t form at all.” If so, one consequence is that “In essence, physicists have been arguing over a trick question for 40 years.” Their claim is controversial at this time. Critics point to other observations which support the “traditional” black hole explanation. What all might agree on is that the new observations and theories show that the universe is, indeed, getting stranger.
1Paolo Coppi, “Inside a Cosmic Train Wreck,” Science, 29 June 2007: Vol. 316. no. 5833, pp. 1852-1854, DOI: 10.1126/science.1139057.
The point of this entry is not to take a position on controversies about star formation, black holes or galactic mergers, but to illustrate the difference between real objects and scientific objects. A scientific object is something about which we cannot know directly through experience: a black hole, a quark, the core of the earth, the interior of the sun, a universal common ancestor, a prebiotic soup, etc. Nobody denies that cars exist, and that if you drive one into a telephone pole, bad things will happen. But scientific objects can only be inferred indirectly. Scientists conceive of their objects as useful entities in equations, and elements of their models in theories. How real are they? That is an entirely different question.
Here we have seen astronomers and cosmologists struggling with and arguing over some scientific objects. There is no question that they “feel” these things are real, and “believe” they are discussing objective reality, but how can they justify those beliefs? As with Darwinism, new and better observations frequently raise new puzzles and occasionally threaten to overthrow what was formerly thought to be well understood. As “elegant” as some ideas may seem, that alone does not prove they represent reality. The universe has no obligation to submit to human measures of elegance.
It may have been elegant to envision galaxies aging slowly, with star formation occurring at a relaxed rate over billions of years. It may have been elegant to envision ellipticals as relics of mergers that stripped away their gas and left them as museums of already-formed stars. Now what? The new observations led Coppi to admit, “While elegant, this story seems wrong.” Now he has to tweak his scientific objects. Now he has to envision a new mechanism that “rapidly and permanently stops star formation,” or has to tweak the models to include feedback from gravitational collapse, or has to keep black holes from colliding. Then Krauss et al come along and claimed black holes are not real. At what point can they claim their scientific objects are real objects?
Dr. Steven Goldman (Lehigh U) produced an interesting 12-hour series for the Teaching Company on this problem: “Science Wars: What Scientists Know and How They Know It.” We’ve mentioned the applicability of these lectures before to questions we often discuss here. In excruciating detail, Goldman gives example after example of controversy in all areas of science for over 2,000 years. Are scientists talking about truth and reality, or are they merely playing games, like members of a fraternity? Do the scientific objects they talk about represent reality or not?
Goldman leaves the controversy open. His only suggestion, offered as a personal opinion in the last lecture, was that we don’t talk about scientific objects as realities, but as actualities – useful entities that allow scientists to make headway in their attempts to understand nature. Yet it should be clear with a little analysis that this is mere quibbling over definitions. Unless an actuality corresponds to reality, what is it? If it isn’t real, or cannot be demonstrated to be real, then what kind of work are scientists doing? That leads to other serious and troubling questions: should the public pay for it? If all they are doing is speculating about things they cannot know, then what value does it have over other kinds of inquiry, that we should grant it epistemic authority and millions of dollars in funding?
Goldman illustrates the point that almost everything scientists thought they knew at the turn of the 20th century is now considered to be wrong. There is hardly any scientific object, whether the earth, the atom, the universe, mass, time, space, the mind, consciousness, or just about anything else from physics to economics, that is looked at the same way today. A logical corollary is that we have no confidence in 2007 that we understand scientific objects so well that our ideas will not be overturned a hundred years hence.
These kinds of questions need to be considered every time scientists talk about the objects of their study as if they are arriving at “the truth” about the universe. Better data, better equipment, and better observations are essential. We are not the ones to judge, however, the point at which our data are so good, and our ideas so solid, that no further scrutiny is needed. The history of scientific revolutions warns us that even Newtonian physics, the epitome of rock-solid science, was vulnerable. This is not to say that we must doubt everything. Rocket scientists, after all, do get spaceships to Saturn at the right spot and the right time. Scientists must be doing something right. When observations continue to contradict theory for decades, though, and when the scientific objects involved are especially remote and far from experience, there is one law that actually gains credibility: Murphy’s.