GUEST POST: Robert Hazen on whether there’s a “missing law of nature”

Plus a stack of notes from me, explaining how Hazen's invigorating ideas are entirely compatible with cosmological natural selection: Darwinian evolution of universes supplies that "missing law"

INTRODUCTION: This is a (heavily annotated!) guest post by one of my favourite modern scientists, the mineralogist and astrobiologist Robert Hazen. I love him because, on December 6, 2006, at a Christmas party, another of my favourite scientists, the late great biophysicist Harold Morowitz, asked Robert a question: “Were there clay minerals in the Archean?” (The Archeon is the second of the four geologic eons of Earth's history, as things started to cool down after the first, fiery, Hadean era, when even the surface of the Earth was mostly hot liquid rock.) And Robert… didn’t know the answer. And if he had stopped there, he would just be another great scientist, but not one of my favourites.

Instead, it occurred to him, with a shock, that probably no mineralogist knew the answer. But not only that – it then occurred to him that no mineralogist had ever even thought to find out the answer. Not only that – he realised that mineralogists hadn’t thought to ask that question about any mineral. It just wasn’t something you did, in mineralogy. Not only that – he realised you couldn’t really do serious research into the origins of life without these answers.

And so he spent the next year recruiting a crack team of mineralogists, geologists, geobiologists, geochemists, and experts on everything from paleotectonics to metamorphic petrology to meteorites – (Dominic Papineau! Wouter Bleeker! John Ferry! Tim McCoy! Dimitri Sverjensky! Robert Downs! Hexiong Yang!)… and they wrote the classic paper Mineral Evolution, which tells the tale of how our Earth started out as a cloud of dust and gas containing not much more than a dozen minerals, and ended up as it is now, with nearly six thousand different minerals. (For comparison, we know of only a hundred minerals on the Earth’s moon; that’s ‘cause there’s no water, no plate tectonics, no weather, no life.)

Of course, when Robert uses the term “evolution” in that paper, he doesn’t mean Darwinian evolution – rocks don’t reproduce, they transform – but he does mean a rich, directional process leading to increasing complexity, stability, and order – as you will see in his fascinating guest post below.

That paper has been of huge benefit to me, as I try to chart the developmental process of an evolved (in the Darwinian sense) universe – particularly the chemical logic that smoothly leads from geosphere to biosphere on suitable worlds.

More recently, he was co-author of a paper with another of my favourite people, the brilliant young astrobiologist and planetary scientist Michael L. Wong. That paper (by Michael L. Wong, Carol E. Cleland, Daniel Arend Jr., Stuart Bartlett, H. James Cleaves II, Heather Demarest, Anirudh Prabhu, Jonathan I. Lunine, and Robert Hazen), On the roles of function and selection in evolving systems, moved Hazen’s initial ideas on mineral evolution up a level, to create a more general theory of the development of structured complexity and order in our universe. And that paper has had a remarkable impact; it’s already being cited widely in the scientific world, but it’s also being talked about by normal human beings; here’s coverage in The Guardian, Reuters, Salon, Vice…

It has touched a nerve, and sent a twitch through the whole culture – and I’m not at all surprised.

Essentially, Robert Hazen and I are exploring exactly the same phenomenon from different sides. He is operating inside this specific universe, pointing out the way in which it develops, step-by-step, in structured, orderly complexity over time. I am operating outside this specific universe, explaining how the Darwinian evolution of universes can lead to exactly that rich developmental outcome, that self-complexification, which he is describing.  (He would tend to use the word “evolution”, of course, where I say “development”; but, again, he doesn’t mean Darwinian evolution. We are actually agreeing with each other pretty closely here, although the terminology can sometimes make that hard to see.) 

Anyway, we’ve been emailing these ideas back and forth for a couple of years. Recently he sent me a link to a short new video he’d made, which gives the essence of the Wong/Cleland/Arend/Bartlett/Cleaves/Demarest/Prabhu/Lunine/Hazen paper On the roles of function and selection in evolving systems.

“Here is the latest from our end. It's too short (7 minutes) and breezy; lacks nuance, as well as important parts of our arguments. But it's out there for people to bounce against.”

(From Bob’s email.)

He’s too modest; this is a great little explainer for a general audience. And if you like it, you can go and read the full paper.

 I asked him if I could put up the video and its transcript here, as a guest post (with my comments), because I think it’s a remarkably good fit with what I’m trying to do at The Egg and the Rock. It gives a clear view of the same problem (explaining the extreme self-complexification of our universe) from a different angle. He said, sure.

So, here you go. Enjoy.

THE VIDEO:

I recommend you watch this. Only seven minutes, and very nicely edited!  Still, I know lots of you prefer to read, or find it easier to take in information through text rather than video, so the transcript (with a couple of screenshots, to help make some points clearer, plus several extremely lengthy notes from me) is below the video.

THE TRANSCRIPT:

Robert Hazen, sitting in a chair, devilishly handsome, looking you straight in the eye:

I have to make a confession here. I have to be honest. We could be wrong. We could be spectacularly wrong. But it's also possible that science is missing a profound truth about the cosmos. We have these 10 or so laws of nature, only one of which currently has an arrow of time. That's the second law of thermodynamics, the increase in entropy—it's disorder; it's decay.

MY NOTE: Ten laws of nature. Nine of them are reversible. One is not. Nine work just the same forwards as backwards; there is no arrow of time. One, the second law of thermodynamics (or, if you prefer, the law of entropy), only works in a single direction. For that one law, there is an arrow of time. The egg can be dropped on the concrete floor, but it cannot be undropped.

BOB AGAIN:

We all grow old. We all die. But the second law doesn't explain why things evolve; why life emerges from non-life. You look around, and you see flowers bloom and trees blossom and birds sing. It seems like all of those things are counter to the idea of disorder. In fact, it's a kind of ordering of nature.

So let me tell you what we think: We think there's a missing law, a second arrow of time that describes this increase in order, and we think it has to do with an increase in information. So there are two possibilities. We could just be wrong. We could be terribly wrong, dramatically wrong. But I think, if we're wrong, we're wrong in a very interesting way. And I think, if we're right, it's profoundly important. 

MY NOTE: I don’t think he’s wrong. But I do think the missing law, the second arrow of time, comes from, or perhaps it’s better to say operates, OUTSIDE our specific universe. It’s the Darwinian evolution of universes – which ultimately, after enough generations, leads to an extremely fine-tuned universe like ours, where the basic parameters of matter have been so finely adjusted by evolution as to lead to a sophisticated developmental process inside the lifetime of this (highly evolved!) universe.

That’s analogous to how the fine-tuning of DNA, over many generations, ultimately leads to a sophisticated developmental process in an egg, so that it develops into a chicken, or a giraffe, or a human being.

And that’s what leads to this extraordinarily unlikely-looking self-complexification – this sprint straight up the cliff of entropy, from hot gas to galaxies to planetary systems to biospheres to intelligent creatures turning matter into technologies – which Bob and his colleagues rightly describe as an increase in information, and then attempt to explain from within our single universe.

But let’s get back to Bob, because he’s about to dig deeper into what he means by evolution, in particular the evolution of minerals – and it’s a tremendously useful insight, which fits in perfectly with my own approach here. I’m pretty sure, by the way, that we are are both right. The Darwinian evolution of universes enables and informs the non-Darwinian evolution of chemicals that he talks about here – because the characteristics of the elements have themselves been formed, and fine-tuned, to do the things he describes, by evolution at the level of universes. (Side note: Holy crap, do we ever need a new vocabulary or agreed terminology of some kind, to keep all these different definitions, and levels, of evolution clear and distinct, so we can understand each other, because at the moment it’s unbelievably messy.) Take it away Bob…

I'm Bob Hazen. I'm a Staff Scientist at the Earth and Planets Laboratory of Carnegie Science in Washington, D.C. I do mineralogy, astrobiology. I love science. We think that, for some reason, there's been a missing second arrow of time. And that arrow has to do with an increase of information, an increase in order, an increase of patterning that goes side by side with the arrow of increasing disorder and increasing chaos, entropy.

The core of everything we've been thinking about, in terms of the missing law, is evolution. When I say the word "evolution," you immediately think of Darwin, but this idea of selection goes much, much beyond Darwin and life. It applies to the evolution of atoms. It applies to the evolution of minerals. It applies to the evolution of planets and atmospheres and oceans. Evolution, which we see as being an increase in diversity, of patterning, in complexity of systems through time.

And so the question is, "Well, what is evolution?" Evolution is simply selection for function. And this applies to every kind of system. Now, in life, you select for organisms that can survive long enough that they can reproduce and have offspring that will pass on their characteristics. That's what Darwin said, and that's one very important example of selection for function. But, in the mineral world, you select for organizations, of assemblies, of structures of atoms that persist, that can last billions of years even in new environments.

They don't break down. They don't dissolve. They don't weather away. It's very analogous to biological evolution, but it's different in detail. We think there's a missing law—it's a law of evolution. And, if there is a law, it has to be quantitative. It has to have a metric. You have to be able to measure something. And what we've zeroed in on is a fascinating concept about information but not just information in general, something called 'functional information.'

Let me see if I can explain this to you 'cause it took me a while to figure it out myself. Imagine a system, an evolving system that has the potential to form vast numbers of different configurations. Let's say they're atoms to make minerals, and you have dozens of different mineral-forming elements, and they can arrange themselves in all different ways. And 99.99999999—I can keep going—percent of those configurations won't work. They will fall apart. They'll never form. A tiny, tiny fraction makes a stable mineral, and you end up with a few stable minerals and lots of rejects.

Now, all you need to do is think about that fraction. If it's one in a hundred trillion, trillion, trillion, trillion possibilities that's stable, then you can represent that fraction as information. And because it's such a tiny, tiny fraction, you need a lot of bits of information to do that—that's functional information. Evolution is simply an increase in functional information because, as you select for better and better outcomes, you select for minerals that are more and more stable. You select for living things that can swim. They can fly. They can see.

You need more information, and each step of the evolutionary ladder leads you to increasing functional information. So, our law, our missing law, the second arrow of time is called the 'Law of increasing functional information.' And that's the parallel arrow of time that we think is out there that we want to understand. 

MY NOTE: I love this, but would, of course, describe it slightly differently – or rather I would interpret what we are seeing here differently. There IS a rise in functional information over time: that rise in functional information over time is of the same kind, and for the same reason, as the rise in functional information in a fertilised egg over time. A Darwinian evolutionary process, outside of, and preceding, the lifetime of the individual egg (or universe) has fine-tuned the genes (or basic parameters of matter) so that their interaction, as time unfolds, will automatically lead to the building out of complexity, and a corresponding rise in functional information.

In each case, you end up with a complex, structured, interacting, system-of-systems, with ever smaller, yet more complex, systems nested inside each other. An egg leads to a body, containing organs, made up of cells, packed with organelles. The Big Bang leads to a universe, containing galaxies, made up of solar systems, packed with biospheres that generate intelligent tool-building creatures. (And if you’re new here, and want to know how those things, those nested levels of increasing complexity, could make a universe more reproductively successful, and could thus have been selected for, read this, or this, or this.)

You can describe it as “finding a law inside the universe”; but it’s more clearly understood as “finding the consequences of a law that has operated across many preceding generations of universe, outside this universe, leading to this observable outcome inside this universe”.

OK, back to Bob!

BOB AGAIN:

The idea of increasing functional information has a really profound implication. Think about the functional information of a coffee cup; you might be holding one right now.

You have a bunch of atoms, and those atoms could be in trillions of trillions of trillions of different configurations, but only a tiny fraction of those configurations will hold a cup of coffee. Now, think about a coffee cup as a paperweight. I know you've used a coffee cup as a paperweight. We all have, and it's pretty good at that, but you can make a better paperweight. And a coffee cup makes a terrible screwdriver. So think about this: We're saying that the coffee cup has value as a coffee cup. It has some value as a paperweight, but it has no value as a screwdriver—that's contextual.

So this is why the second arrow of time is difficult for science because it's saying there's something in the natural world that is not absolute. It's contextual. It depends on what your purpose is. It depends on what your function is. If it's true, what we're saying is there's something in the Universe that is increasing order, it's increasing complexity, and it isn't doing this in a random way. It's selecting for function. And if it is, if you're selecting for function, it means that there almost seems to be—can I use the word "purpose?"

Do minerals have a purpose? Do atmospheres have a purpose? Does life have a purpose? To me, there's something real there, and the old way of thinking of a single arrow of time no longer rings true to me.

–Robert Hazen

FINAL NOTE FROM ME:

OK; that’s the end of the transcript. Let me sum up my thoughts: Bob is totally right. (And Michael Wong, and the whole gang)  “There's something in the Universe that is increasing order, it's increasing complexity, and it isn't doing this in a random way. It's selecting for function.” 

Yeah, there is. AND I KNOW WHAT IT IS.

OK, let me briefly recap the theory I’ve been exploring, and expanding – Cosmological Natural Selection – and show how well it complements, and explains, Bob’s words.

COSMOLOGICAL NATURAL SELECTION, UPDATED: A RECAP

Remember; a black hole is a singularity, a point of maximum compression, where matter has collapsed under its own gravity so fucking hard that it has left its parent universe. Nothing, not even light, is coming back from there. It’s travelled beyond all our current theories; even the mathematics of General Relativity breaks down inside a black hole and starts to vomit infinities. A black hole has bent the spacetime around it so hard, it has budded off. It’s gone. Bye bye.

And remember, each new “child” universe is a brand new spacetime, which comes into existence in a Big Bang, expanding from a singularity – a point of maximum compression. My argument (which is based on Lee Smolin’s argument, which is developed from John Wheeler’s argument) is that an expanding-from-a-singularity Big Bang baby-universe (like ours) is simply the next step, the other side, of the contracting singularity which is a black hole in a “parent” universe. And each such new universe is defined by roughly 27 basic parameters of matter. That’s its rule-set (which you can think of as vaguely analogous to its genes).

The 27 basic parameters of matter in each new child universe can vary very subtly, very slightly, from those in its parent. That variation makes it more or less likely to produce more or less black holes (to be more or less reproductively successful); and so Darwinian evolution inevitably follows. Over enough generations, even if you started with the most simple blob-of-glop universe that could only collapse directly into a couple of black holes, you will eventually (and this process can be completely blind, completely random; this is just the beautiful, and relentlessly bountiful, logic of evolution), eventually fine-tune your way to universes which are far more reproductively successful – which produce HUGE numbers of black holes – and anything that leads to such increasing numbers of black holes will be conserved. Will be selected for. (Sure, the less reproductively successful universes will also still chug away, reproducing less successfully – but there will be, by definition, a lot less of them.)

So the “something” that is increasing order is the extreme fine-tuning of those basic parameters (the strong nuclear force, the weak force, the mass of the electron, and so on). And that fine-tuning was done by Darwinian evolution, acting on ancestral universes, all the way along the evolutionary line that has led to this complex, reproductively successful universe of ours, with its forty quintillion black holes leading to forty quintillion Big Bangs; forty quintillion offspring. (What else could have so thoroughly fine-tuned our specific universe, other than Darwinian evolution? Give me another mechanism that we know of that can do such fine-tuning.)

And so there has already been a huge amount of selecting for function, before the Big Bang; the mass of the electron has been selected for function. The exact strength of the strong nuclear force has also been selected for function. (As has the point at which that force flips from attractive to repulsive.) That’s how and why the elements of the periodic table can be so swiftly and efficiently built in the heart of stars (by fusion), and distributed through supernova explosions (when the strong nuclear force flips from attraction to repulsion, as the star collapses, blowing those new elements back up out of the star’s gravity well). And so on, across all those parameters.

At this point in the long evolutionary history of universes, along our specific universe’s evolutionary line, it’s the function of those parameters to build out stars and galaxies and a full suite of elements and thus planets and biospheres, because those things ultimately lead to the reproductive success of universes, in all the ways I’ve described elsewhere, but I will now describe yet again, because it’s so important, and so interesting. (And because I’m constantly trying to explain it better; this website is where I’m working out the articulation of these ideas for the eventual book, so give me feedback at the end: is this clear? What can be improved?)

HOW THE EVOLUTION OF UNIVERSES HAS LEAD TO REPRODUCTIVE HYPER-SUCCESS: THE THREE KINDS OF BLACK HOLE

This (highly evolved) universe we live in is capable of producing (and does produce) three kinds of black hole. Each kind is the result of an evolutionary breakthrough, not in the lifetime of our specific universe, but in its distant evolutionary past. Each breakthrough built on the preceding breakthroughs, which were conserved; thus our specific universe, shortly after the Big Bang, first produces a modest number of primitive, easy-to-make, direct collapse supermassive black holes. (All universes that can reproduce have always been able to do that, going all the way back; it’s the simplest form of reproduction of universes, requiring no complex structures or even elements.)

Those supermassive black holes help build out galaxies of stars around themselves in ways I’ve described in detail elsewhere… which eventually generates a far larger number of more-complex-to-make, medium-sized, stellar-mass black holes, as the larger stars run out of fuel, collapse, and explode.

But once those stars have built the periodic table (by fusion) and distributed it (by exploding), many new stars (like our own) form from that enriched gas and dust, along with planets (and their moons) on which intelligent life develops, and then (to efficiently power itself) generates a huge number of much smaller, incredibly-complex-to-make, technologically-built black holes. (Because black holes convert mass into energy fifty times more efficiently than fusion.)

(A quick history lesson, to make sense of this: The original version of cosmological natural selection, by Lee Smolin, in the 1990s, just assumed our universe produced one kind of black hole – stellar-mass/stellar collapse black holes – with anything larger simply being made from a bunch of those merging; this made for an interesting, but speculative and rather static theory that didn’t explain life, or supermassive black holes. Over the next decade or two, Clément Vidal, John Smart, Michael E. Price, Edward R. Harrison, Louis Crane and others extended the theory to incorporate intelligence, and the intelligent technological manufacture of vast numbers of small black holes (for maximally efficient energy production), thus explaining why life might have evolved, and been conserved – it made for much more reproductively successful universes. And in the past few years, I have extended the theory again, by pointing out that the logic of evolution implies that the supermassive black holes found at the heart of galaxies are almost certainly an evolutionarily conserved, extremely primitive form of reproduction that goes right back to the earliest universes, and that such supermassive black holes should therefore form shortly after the Big Bang by direct collapse, not later by merger of smaller stellar mass black holes. That is, supermassive black holes should form first, and generate galaxies of stars around themselves. This is what the James Webb Space Telescope subsequently found.)

You will note that each stage – direct collapse supermassive black hole; stellar collapse (medium sized) black hole; technologically produced (small) black hole – requires more of the periodic table – more and heavier and more complex elements – to produce smaller (and thus more numerous, and thus more reproductively successful for the universe) black holes; that’s because each stage came later in the evolutionary history of universes; the periodic table, too, has evolved and complexified.

And it’s all that hidden-from-view selection of function, which preceded the Big Bang, which allows for all the selection for function inside this universe.

But what a liberating, glorious thought that is, that there is so much selection for function still going on inside this universe! Because that means evolution is most certainly not over. Universes have not been perfected.  And above all, that means this universe isn’t predetermined: matter is swiftly and efficiently built and distributed, but that matter nonetheless then has to explore the possibility space of this brand-new universe, ever-so-slightly different from its parent. Minerals will complexify wherever they can – on the surface of wet rocky planets like Earth, sure, but also deep in the liquid water oceans of icy moons with molten cores. Different worlds, under their very different conditions, will generate different minerals. It is not guaranteed you will get calcite, say, or amphibole, specifically. But you will get some complex, stable minerals, appropriate to their environment. And those geospheres will tend over time to become biospheres. (Because that has been selected for! That’s what carbon and oxygen and hydrogen and nitrogen and sulphur and phosphorus and sodium and potassium, in this kind of highly evolved universe, do.)

At which point it is not guaranteed you will get human beings – that apes will become smart. Could be something more like an octopus, or a bear, or something really weird. Depends on the conditions of that particular planet, that particular icy moon. But something will become smart, and start to manipulate matter in ways which Nature alone, without life, cannot. Ways which will inevitably lead that smart something to the more and more efficient production of energy, to power whatever it chooses to do, and thus ultimately leads to the technological production of small black holes. (Which is why evolution, at the level of universes, selects so hard for intelligent, technology-wielding life, and evolution allows it to take over worlds so fast once it has emerged; it leads to massive reproductive success for that universe.) And right here, right now, that something happens to be us: We are the growing tip of a living universe, transforming it; bringing into being things never seen before. We are evolution in action. We are the agents, the 0.0000000000000000000001% of matter in this universe with agency, generating and processing ever more functional information, looking up at the planets, looking up at the stars, and thinking; what can I make this into? What do I want this universe to become?

HOLY SHIT, IT IS AMAZING TO BE US, RIGHT HERE, RIGHT NOW…

There’s a real terror in the general scientific community about this kind of talk, by the way. A terror that you are about to summon God as an explanation for purpose, for direction, for order. And they are partly right to be terrified, because “God”, as usually summonsed by believers, isn’t a useful explanation, it’s an anti-scientific dead-end; great, you’ve “explained” the universe – “God made it!” – but all the weary work of explanation now has to begin all over again, to explain how the hell this new-level-of-reality God-thing (that we can’t actually see) came about, and got so complicated. Nothing has been explained.

But that is why cosmological natural selection, or the evolved universe hypothesis, is so beautiful, and powerful. It allows you to acknowledge, in full, the deep weird beauty of a universe that gets richer and more complex over time, that generates more and more structured order until it is nothing but gushers of life orbiting fountains of light circling vast spirals of fresh star production; and it explains all that, it gives you a full explanatory mechanism, without having to add anything new at all to our existing science, and without having to subtract anything. No made-up shit, no woo-woo. It just applies Darwin to universes, and clarifies everything.

IMAGINE NO RELIGION

But there is something here that is sympathetic to the impulse behind religion, even if it isn’t sympathetic to the concretised beliefs of any narrowly specific religion. Such a universe – a highly evolved universe selected for function, travelling along a developmental path where each level in turn selects for function – has many of the attributes human beings have traditionally associated with something human beings have usually called “God”.

After Darwin, evolution (based on inherited variations in genes) came to explain the complex, orderly structures of life, and how they came into being, step by step, generation by generation, from an earlier absolute chemical simplicity. Likewise, we are starting to see that evolution (based on inherited variations in the basic parameters of matter) best explains the complex, orderly structures of our universe, and how such a universe could come into being, step by step, generation by generation, from an earlier absolute simplicity in the fundamentals of matter itself. In both cases, evolution takes the place of God, without in any way detracting from the wonder and glory of the thing being explained.

And so here we are, in a gloriously ironic, and paradoxical, situation; with mainstream science denying clear evidence of evolution, in the mistaken belief that they are denying evidence of God.

OK, I’m done. Did I talk all over Bob? I did talk all over Bob. But hey, that’s the value I can add here. If you want to read him again, uninterrupted, there’s a clean transcript over on Big Think, from whence Bob let me borrow it.

And let me say again, it’s worth reading the original paper in full. There’s a lot of useful nuance in the paper that was perforce left out of this brief popular overview.

But basically, we are saying the same thing:

1. The way this universe channels its energy, so as to produce ever more complex, self-sustaining, self-complexifying structures, in the teeth of the second law of thermodynamics, requires an explanation. 

2. This universe clearly has purpose. Its parts have purpose. The complexification of those parts has a purpose. It is going somewhere and it is doing something and we are a vital part of that going, that doing.

3. There is an explanation for all this, and it is totally compatible with science.

Here is where we differ: Bob thinks the explanation is a law of increased functional information.

I think that an increase in functional information is indeed an aspect of our specific universe, but that it isn’t exactly a “law”, as such; it’s best understood as a downstream consequence of the fine-tuning of the basic parameters of matter by Darwinian evolution at the level of universes.

OK! We’re done. Share this with anyone you think might be interested. And throw in your thoughts in the comments.