Predictions! What the James Webb Space Telescope will see (and not see)...
Galaxy formation, direct collapse black holes, and why quasars do what they do
This is not a normal post, and will be harder going for the general reader than my usual stuff (and much less fun), so I strongly advise you to just stop if you are not enjoying it.
It’s not you, it’s me: I’m a slow writer, and this area is very hard to write about for a broad audience, because so much background information – theory, observation, history – has to be delivered to provide context so that the predictions make some sense. As a result, my original predictions post for a general audience has grown too long and unwieldy, and I’ve had to break it into several posts, which I am still working on.
But the James Webb Space Telescope delivers its first data in four days, on July 12th.
So I am in danger of delivering several immaculately written predictions… after the relevant data is revealed.
That is not good.
I am therefore putting up this extremely minimal, and unedited, post-with-no-jokes today, simply to register my predictions in public. I will, I hope, fill them out over the next few days and weeks with context and background and all the stuff needed for the general reader to understand and enjoy them.
But today’s post is mostly just a bunch of terse bullet points, to get some of the raw predictions into the public domain in time.
THEORETICAL BACKGROUND TO THE PREDICTIONS
I’m going to assume (because I think the evidence is now overwhelming) that our universe is the result of an evolutionary process at the level of universes. As a working model of the possible mechanism, I’m going with Lee Smolin’s theory of cosmological natural selection: black holes in parent universes give birth to Big Bangs in child universes. (Singularities “bounce”, as the great John Wheeler argued.) The basic parameters of matter in the new, baby, universe are very slightly altered from those in the old, parent, universe, in each generation. This leads to different rates of reproductive success (ie, less or more black hole production). And that leads, inevitably, to evolution.
Thus, my baseline assumption, from which I am making these predictions, is that our universe evolved, is from a reproductively successful line of universes, and is therefore optimised for black hole production over the course of its lifetime. (In evolutionary terms, it’s close to, or at, a local fitness peak for black hole production, given the basic parameters of matter it inherited.)
GENERAL BROAD-BRUSH PREDICTION
As a result of that evolutionary optimisation of its parameters…
Our specific universe develops, over time, in a way analogous to a complex biological system. (With the important difference that a universe is both organism and environment, and therefore must generate, through gravitational collapse and fusion, the energy that it consumes.)
Throughout the unfolding lifetime of our universe, then, each developmental step drives the next, in a very specific order, that has an evolved logic, and is driven by the complex interaction of the evolved parameters at both large and small scales.
Just as with an evolved biological organism, the energy that moves through the system organises the system.
SLIGHTLY MORE SPECIFIC PREDICTIONS
Here we go…
THE ERA IMMEDIATELY AFTER RECOMBINATION IS OPTIMISED FOR SUPERMASSIVE BLACK HOLE FORMATION
I agree with mainstream cosmology that, about 400,000 years after the Big Bang (at the point of recombination, where protons and electrons cooled down enough to recombine and form neutral hydrogen and helium), we have a smooth, featureless, expanding, hydrogen-and-helium-only universe. We can see from the Cosmic Microwave Background Radiation that this gas contains no significant density differentials. It also has no catalysts (no carbon, no oxygen, etc) for cooling, fragmentation, and thus star production.
My argument is that this situation has been optimised (by evolution at the level of universes) for supermassive black hole formation. The early universe, after 400,000 years (post recombination), is analogous to a supersaturated solution in an extremely pure liquid, where you can dissolve more of a solid than the liquid can normally hold, because there are no seeds to trigger the process of crystallisation. But once crystallisation IS triggered, and DOES begin, a huge amount of crystallisation immediately happens throughout the entire liquid. (That is just an analogy! But it gets across the general idea.)
And so we will discover that huge numbers of supermassive black holes form directly from this gas, with immense areas of gas collapsing without fragmentation (ie, without the collapsing cloud breaking up into smaller parts which form stars – because there’s no carbon or oxygen to form the CO, O2 and H20 to catalyse the necessary cooling and fragmentation). This is basically the model of direct collapse black holes theorised by Volker Bromm, Abraham Loeb, Priyamvada Natarajan, Fabio Pacucci, Andrea Ferrara, and others.
(Side prediction: at least one of those, Pacucci, Bromm, Natarajan, Ferrara, and Loeb – and very likely more than one – will eventually win the Nobel Prize for their work on direct collapse black holes.)
THE BRIEF ERA OF DIRECT COLLAPSE BLACK HOLE FORMATION
The era of supermassive black hole formation will turn out to be very specific, well-defined, and short. At the absolute latest, supermassive black hole formation will be over, throughout the universe, within 300 million years of the Big Bang (that’s roughly 2% of its current age), but I suspect much, much earlier. See “The brief era of direct collapse black hole formation” (by Bin Yue, Andrea Ferrara, Ruben Salvaterra, Yidong Xu, and Xuelei Chen).
These huge numbers of supermassive black holes across the universe almost immediately attract gas, fire up quasars, and begin (I argue) optimising the surrounding area for star formation. This is best seen as the complex developmental process of an evolved entity.
Bear in mind, everything that makes it easy to make supermassive black holes (no carbon, no oxygen, no density differentials, etc) makes it incredibly hard to make stars.
So those supermassive black holes now need to completely change the environment, and optimise it for star formation. Which of course, ultimately, means small-black-hole formation, (because small black holes are made by old stars collapsing), and thus reproductive success.
Oh, by the way, there won't be a huge range of sizes for those primordial black holes: they will have been optimised (by evolution, at the level of universes) for galaxy formation, and will therefore all be in the supermassive range; the vast majority of them will have the mass of at least hundreds of thousands of suns (many of them almost certainly far, far more), right from the start. This is the Shock and Awe approach to galaxy formation, which evolution is highly likely to favour, as it is more efficient. Obviously, as they pull in gas, they will grow; but they START OFF REALLY FUCKING BIG, many tens of thousands of times bigger than your average stellar collapse black hole.
(It is possible that black holes in another highly specific size range – orders of magnitude smaller – will also be created for another purpose; but what we will NOT see is a random distribution of primordial black hole sizes).
Those primordial supermassive black holes (with as much mass as at least hundreds of thousands of suns) will form the gravitational seeds that will collapse the early clouds of otherwise evenly-distributed gas, leading to the formation of galaxies.
HERE COME THE WARM JETS (QUASARS)
That means quasars, too, will switch on before stars form.
Remember, quasars are what happens when gas is pulled towards a supermassive black hole by its immense gravity. That gas moves closer, and spins ever-faster, in the huge magnetic field around the spinning black hole. Because the gas, being ionised, is electrically charged, that generates an even huger magnetic field, and the gas therefore becomes contained inside a blazing hot magnetised donut – rather like the Tokamak fusion reactors we have tried to build on earth. The magnetic donut then pinches ionised particles incredibly tight, before blasting them out of the holes in the donut, at a significant percentage of the speed of light, in collimated jets (ie, the particles shoot away all lined up and parallel, in a narrow beam – just like the particles shot from a particle accelerator here on earth). It’s optimised, efficient… and takes insane amounts of energy.
WAIT, HANG ON A MINUTE, WHERE DOES ALL THAT ENERGY COME FROM?
The astounding quantity of energy required to do all this is taken from the spin of the black hole by either the Blandford–Znajek process or the Penrose mechanism (click the links for details), both of which are ASTONISHINGLY efficient. (As you would expect of a vital organ in an evolved universe performing a crucial developmental task.) The conversion of matter to energy by these processes is far far more efficient than even the efficiency of nuclear fusion. (30 or even 40% of matter can be converted into energy by gravitational collapse into a black hole, whereas fusion in a star converts less than .7%)
Mainstream reductionist science kind of shrugs at this clear example of hyper-optimised efficiency, because, under the current paradigm, nothing is allowed to mean anything.
An evolved universe theory, though, says anything that’s that optimised and efficient is probably doing an important job in the development of the evolved organism that is the universe. And this is the job I argue it is doing…
HOW QUASARS WILL ACCELERATE STAR FORMATION
The quasar will massively accelerate galaxy formation in four ways:
1. FUSION AND DISTRIBUTION OF CARBON
Its insanely hot and powerful jets will fuse light elements (mostly carbon and some oxygen), and distribute them back out into the gas cloud.
This is where I am taking a bit of a risk in my predictions. We already know quasars can fuse and distribute carbon and oxygen – see, for instance, the paper Light-Element Nucleosynthesis From Jet-Cloud Interactions in Active Galactic Nuclei.
I’m predicting that in the early universe, that process is vital for rapid galaxy formation; it precedes, and enables, large-scale star formation in galaxies; is far more intense than we see in the more recent universe; and is widespread.
Also, we have a strange excess of carbon in many very early gas clouds: I am arguing that this is where it comes from, rather than from classic stellar nucleosynthesis, distributed by supernova. Obviously classic stellar nucleosynthesis takes over carbon production later, but I argue that the very first carbon is made and distributed by quasars to facilitate the earliest star (and thus galaxy) formation.
OK, second way in which quasars facilitate galaxy formation. This one is less controversial; the mainstream would largely agree with me here, although they would see it as accidental and meaningless, and I would see it as more evidence of optimisation by evolution of the process of development.
2. REMOVING ANGULAR MOMENTUM
The jets blasting from the quasar will also efficiently remove an absolutely stunning amount of angular momentum from the collapsing gas cloud, allowing it to collapse faster. This faster collapse will both feed the quasar faster, AND let stars form faster. Efficient!
And on to the third way quasars help form galaxies…
3. PRESSURE WAVES, CAUSING STAR FORMATION
The quasar’s jets will also send pressure waves through the gas cloud. Like a barista banging a jug of steamed milk off the counter, to collapse all the big bubbles, those pressure waves will collapse the freshly enriched gas into waves of star formation.
As a result, galaxies will form efficiently, and early. I argue that huge numbers of supermassive black holes and their quasars will be blazing away merrily, well inside the first fifty million years. There will be absolutely loads of recognisable, rapidly growing (rapidly star-forming) galaxies within the first 100 million years (probably much sooner). This is earlier than the mainstream have traditionally assumed. (They keep having to shuffle a bit further back, as they find new quasars, and their galaxies, ever further back in time. But they are pushed there, reluctantly, against the logic of their paradigm; I am leaping there, exultantly, because my paradigm predicts it.)
So the James Webb Space Telescope will basically see galaxies with active galactic nuclei (ie, quasars and jets) all the way back, because those active nuclei come first and are what form galaxies.
Quasars are fairly rare now; I am arguing that they will turn out to be ubiquitous in the very earliest stages of the universe because they are the engines that build out galaxies.
Let me make two really specific predictions, to firm this up:
REALLY SPECIFIC PREDICTIONS
Both HD-1 and CR-7 (click on them if you want to know more… and yes, CR-7 was knowingly named after Christian Ronaldo – scientists are human beings!) will turn out to be active galactic nuclei: that is, supermassive black holes and quasars – and not simply clusters of (hot, young, hydrogen-and-helium-only) Population III Stars. (Yes, for irritating historical reasons, astronomers call these imaginary, theoretical, first-ever stars Population III stars. Yes, this is a terrible name for the first-ever stars. Astronomers have many virtues, but none of them are literary. If they ever wrote a novel, the hero would be called something like Protagonist 6b, and be introduced on the second-last page.)
Oh and they found a lot of carbon in the spectrum of CR-7 back in 2018: I’m arguing it is made, and distributed, mostly by the quasar.
Oh, same goes for GN-z11. It will turn out to be a supermassive black hole and quasar, driving star formation.
WHAT IS NEW ABOUT THESE PREDICTIONS; GIVEN THAT THEY ARE LARGELY BASED ON THE EXISTING SCIENCE?
Well, on quasars, I’m pretty much saying the opposite to the mainstream. The mainstream now accuse quasars of quenching star formation by heating the gas, and blasting gas out of the galaxies. This is a classic “it’s-all-random”, reductionist bullshit attitude. My argument is that active galactic nuclei, and quasars in particular, are in fact optimised to facilitate star formation: the reason you see star formation end early in many galaxies with active galactic nuclei isn't because big-ass quasars shooting out relativistic jets at near lightspeed wreck the process of (random) star formation, but entirely the opposite: those jets have enabled, and accelerated, star formation so thoroughly that the majority of the available gas has been quickly turned into stars. (It’s just that we missed seeing most of this process, because it happened in the first billion years, when we can’t see shit.) Having efficiently done their job, the quasars switch off. The gas is now stars…
Because we can often only see this final part of the process, we get the causality confused, and think the jets killed the star formation. But it’s simply that the quasar’s job is done: the galaxy is built. Those stars are now capable of making more stars on their own. Supernovas are distributing newly enriched gas, and making black holes; win/win! The spiral arms are in place; permanent rotating pressure waves, perpetually channeling gas along magnetic field lines to their star-making regions, to make fresh stars.
That means a spiral galaxy is now self-sufficient. Power down the engine, job done.
OK, let’s throw in two corollaries, and a caveat, before we wrap this post up.
THERE WILL BE NO ERA TOTALLY DOMINATED BY POPULATION III STARS
Remember, I am saying supermassive black holes form in a relatively short, sharp period just after recombination – before there is widespread star formation – and that they pretty much immediately generate quasars, which fuse and distribute carbon into the gas clouds from which most of the first stars are formed. So there will be no era totally dominated by so-called Population III stars – the hypothesised (highly inefficient) first stars, containing only hydrogen and helium.
Most of the first generation of stars will, if I am right, contain traces of carbon at formation, because early quasars make it by fusion and distribute it into the clouds to seed star formation. And such stars will therefore be relatively efficient at fusion, element formation, etc. (They will still be very low in carbon, and other elements such as oxygen, relative to later stars; but not completely lacking, as Population III stars are theorised to be.)
Minor possible exception: Priyamvada Natarajan argues that a certain number of Population III stars are needed, in order to trigger the direct collapse of supermassive black holes.
Her argument is that in order for the gas to collapse totally smoothly and without fragmentation, it can’t contain H2 molecules. (H2 is just two hydrogen atoms, bonded together.) This is because H2 molecules would cool the gas cloud locally, and thus cause fragmentation, and thus star formation – which would prevent the huge smooth supermassive collapse of the entire area into a black hole. So she argues that soft ultraviolet light from fast-burning Population III stars breaks up H2 molecules in the neighbouring smooth cloud, without actually ionising the hydrogen. That enables a smooth collapse.
I LIKE THIS THEORY (POPULATION III STARS AS SPARK PLUGS)
I like this theory, because it means Population III stars aren’t totally inefficient, short-lived and useless stars (as they are in most theories at present), but are in fact highly optimised to perform a specific function: they produce a huge amount of ultraviolet light very fast and early, at just the right frequencies to trigger supermassive black hole collapse.
They are the spark plugs providing the vital spark that triggers supermassive black hole formation.
So, she might be right, in which case we will see some Population III stars. But I would argue, not a huge number. They’re the spark plugs, not the engine.
It is also possible that the basic parameters of matter are simply so finely adjusted by evolution as to enable a huge wave of spontaneous collapse into supermassive black holes, without any spark; that it’s just a phase transition that happens after recombination, inevitable, baked into the nature of matter. I would like this, as it would be very clean: black holes would be primary, and stars secondary. That makes sense, from an evolved universe point of view, as all prior universes must have generated black holes (in order to have reproduced); but they don’t need to have generated stars.
Well, we will see. Bear in mind, we have never seen a single, actual, hydrogen-and-helium-only Population III star so far, so we are all groping around in the dark here until the James Webb Space Telescope sends us some pictures.
WE WILL DISCOVER THAT OUR UNIVERSE WAS NEVER DOMINATED BY SMALL GALAXIES
Bear in mind that observations so far (pre-James Webb) have shown us that there are somewhere from 90% to 99% fewer small galaxies than there should be, if our current dark-matter-based theories of galaxy formation are true.
An evolutionary theory gives an explanation for this.
If primordial supermassive black holes (and their quasars and jets) seeded galaxy formation very early on, then the number of large galaxies in the early universe will be large, compared to the number predicted by the current, mainstream LCDM model. (LCDM, or ΛCDM, stands for Lambda/Cold Dark Matter, where Lambda/Λ represents dark energy.)
And, contrariwise, the number of small galaxies we see in the early universe (the first billion years, say) will be very small, relative to the number predicted by the current, mainstream LCDM (Dark Energy/Cold Dark Matter) model.
This is because the universe didn't start off – as our current models assume – with a fairly random cloud of gas, from which the first stars slowly formed (with the vague and accidental help of clouds of dark matter)… then attracted each other gravitationally to form small clusters… which merged to form small galaxies… which then attracted each other and joined together to form large galaxies… which began to develop larger and larger galactic cores. (A bottom-up theory of galaxy formation.)
In my version of an evolved universe model, galactic cores exist, as supermassive black holes, and, almost immediately afterwards, quasars, before the first stars form in any significant numbers.
That is, an evolved universe theory would argue that large galaxies start to develop from day one: Any random wisp of gas will therefore have been far more likely to be attracted to the nearest supermassive black hole than to any small local minor density fluctuation.
(If you were a cloud of gas forced by gravity to choose between a nearby, minor, density fluctuation to your left, and a black hole with the mass of many million suns, a little further away, to your right… well, which way are you going to fall?)
Are there going to be some small galaxies early on? Sure. This is an evolved process, organic and messy, it's not mechanical and perfect. But there will be WAY fewer small galaxies – at least an order of magnitude fewer – than current models of early galaxy formation predict.
And of course smaller galaxies will merge to form larger galaxies; but no matter how far back you go, no matter how early, most of those “smaller” galaxies will be pretty big, compared to those predicted by our current models.
AND A CAVEAT
This one is a little cheeky, but let’s go for it. I think there will turn out to be two slightly different galaxy formation mechanisms, one for ellipticals (simple, blobby galaxies), and another for spirals (complex, structured galaxies). Oh man, that’s definitely going to need its own post… But anyway, in short:
Ellipticals will turn out to form very fast, very early, and very efficiently, through a very simple quasar-driven process. That process uses up all the gas in their immediate region quickly and efficiently, after which the elliptical galaxy can't make any more stars.
Spirals, however, will build out a more complex structure (both physical and electromagnetic), that allows them to channel gas from outside their immediate region into their star-making regions. That process involves driving ionised gas along electromagnetic field lines.
In simple terms, elliptical galaxies just sit there, eat the gas in their immediate vicinity, and process it into stars (and, later, black holes). They grow VERY fast, and stop growing young, when they run out of food.
Spiral galaxies, however, can hunt their food (clouds of intergalactic gas, to make fresh stars), by driving it along magnetic filaments, and therefore can keep on growing (while maintaining their complex structure) for many billions of years.
My guess is that these two types of galaxy are from two very different eras in the evolution of universes.
AN ANALOGY FROM BIOLOGY
An analogy would be with the way DNA evolution, inside the lifetime of this universe, has left us today with both prokaryotic cells (simple blobby cells – think E. coli), and eukaryotic cells (complex, structured cells of the kind you find in modern plants and animals), even though they evolved at very different times.
That’s because evolution conserves successful organs, and organisms…
And yes I know that the mainstream theory for decades was that elliptical galaxies were made when two spiral galaxies collided. I strongly suspect that theory will turn out to be wrong. It’s like saying when two gazelles collide, they make a tortoise.
Bear in mind, all our guesses about what happens when two galaxies collide are based on snapshots. We can’t see the actual process unfold; we can just see the frozen state of play in a large number of different collisions, and try to assemble a coherent story out of these different moments.
I feel strongly that in an evolved universe, spiral galaxies will prove to be surprisingly coherent and robust. I think we are only beginning to understand the extent to which spiral galaxies structure themselves electromagnetically, as well as gravitationally. When two spiral galaxies collide, most of the time, I suspect, they will merge to form a bigger spiral galaxy, their individual complex magnetic structures will merge to form a single, new, complex magnetic spiral structure.
OK, that will do for now. A ragged post, but it has the virtue of existing, and of being posted before the James Webb data arrives.
Please give me feedback, if you see any glaring errors! And well done getting to the end of a post with so few jokes.
Yes, please do pass this on to anyone you think would be interested…
Oh, and if you are not yet a subscriber, but you found this interesting, please do subscribe. (Don’t worry, most posts contain more jokes.)
Special thanks to everyone who read drafts of this and/or earlier posts and gave feedback, in particular the evolutionary systems biologist and philosopher Johannes Jaeger, who hugely improved the evolutionary aspects; Kelvin Long (physicist, aerospace engineer, and extremely helpful editor of the Journal of the British Interplanetary Society) who defended the honour of reductionism, and clarified my thinking greatly with his thoughtful feedback; PJ King (an actual rocket scientist – he has designed rockets that flew!), who saved me from several grievous errors; and Solana Joy and Sophie Helga Gough Fives, who both helped greatly with structure and general edits. All remaining errors and idiocies are entirely my own fault.
It’s like saying when two gazelles collide, they make a tortoise.
Great clarity and damn funny.
I really liked this post. If I'm not mistaken, the recent news where JWST has seen many more mature galaxies in the early universe is a point in favor of the model you write about here.
You're a hard man to reach! My first e-mail to you got caught in the flood of new subscribers (congratulations, it's well deserved). The second was an automatic reply about the mailbox being unmonitored. I guess you got too many e-mails with all the subscribers. Anyway, I'm pretty sure the universe intended for us to talk so I made a Substack for it: https://plantser.substack.com/p/hi-julian-im-micke