Original Post (with comments)
(Warning – this is longer than usual. What’d you think – enlightenment is free?)
How do you get from primordial soup to living cell? That’s it. That’s the kryptonite in the creationist’s napsack. But that seems to be where it stays. They rarely (if ever) actually pull it out and examine it. How many times I have heard, “Ah, but intelligent design disproves everything you’re saying.” When I try to respond, deaf ears. The answer is somewhat complicated, so I somewhat understand, but it somewhat irks me that many creationists aren’t willing to fully examine what they believe, especially given how vehemently they believe them. This I also understand…somewhat.
They assume the Kryptonite works because it was handed to them by someone in whom they have implicit trust. If they put it to the test, this, the last hope for an argument that has been all but decimated in every debate it has entered, what happens if it fails? What happens if the evolutionary scientist is not weakened by the moment? What happens if the non-believer walks up, grabs the rock, crushes it into a fine powder, and sniffs some of it up his nose, and lives?! These are serious questions for some. Not me.
One thing I like about believing as I do is that I really have nothing invested in my beliefs, with the exception of my admittedly irrational belief in rationality as a superior method of thinking. But beyond that, I could change my mind about anything. Sure, it might be hard to get used to something new, but I’d be OK. All I need is for an assertion to meet my evidentiary requirements, then I’m the first to start exploring the logical consequences of it on my life. Creationists, however, do not enjoy such luxuries. Their belief in God’s creation of the world is the capstone on the arch of their religious faith. If it falls, the arch falls, and life as they know it changes forever. As I am the type to rip off the band-aid all at once, to the truly open-minded, I say get on with it. So let’s pull out the Kryptonite and take a look.
Here’s what’s about to happen, and it may not be pretty. I’m going to put forward a theory for how you get from non-living chemicals floating around in a liquid to a living cell. It all comes down to chemistry and the theory of self-organization, which is most eloquently articulated by Stu Kauffman in, At Home In The Universe. (You simply cannot consider yourself a scientist without knowing what’s inside this book.) I will start by drawing attention to the fact that something this ostensibly miraculous is actually not so uncommon in nature – we’ll look at phase transitions. Then, we’ll look at how connections between entities can become so complex that the entity itself eventually becomes something entirely different (via, you guessed it, a phase transition). I’ll then bring these two ideas together by talking about a chemical reaction network and how you to get to catalytic closure. Finally, we’ll use some back of the envelope stats to conclude that life from chemicals in a soup is not only possible; it’s probable. So consider yourself warned – serious scientifigeekification ahead.
The idea that something seemingly random can suddenly transform into something orderly may seem strange. In the world of science, these transformations are called phase transitions. The easiest example is ice. The arrangement of the molecules in liquid water is pretty much random in the sense that the nature of water is not dependent upon the specific arrangement of water molecules. But suddenly, when the temperature of the water reaches 0 degrees Celsius, the water molecules assemble themselves into the orderly substance we call ice. If you look at ice molecules under a microscope, they are crystallized in a stacked arrangement – an orderly arrangement. Ice is, in fact, defined by the arrangement of the constituent molecules as much as it is by the temperature. The phase transition in this case is the change in the physical state from liquid to solid, which corresponds to a change from disorder to order.
The point of this is to suggest that the emergence of complex life was just a phase transition, where a teeming soup of replicating molecules transformed into a cornucopia of living diversity. It’s really about connections. The following toy problem illustrates what I mean.
Buttons and Thread
Imagine throwing 1000 buttons onto a hardwood floor. Now randomly pick two buttons and connect them with a string of thread, and put them back down. Keep doing this. Sometimes you’ll pick up two buttons you haven’t picked up before. Sometimes, one or the other will already have a thread tied to it. No matter, you just keep connecting buttons. Over time, you’ll find that when you pick up some buttons, they are interconnected with several others. This is an example of a random graph. We’ll call these interconnected clusters webs. If you keep glancing at the whole floor as you connect pairs of buttons, you’ll start to see that more and more webs are emerging. You’ll start to see “islands” of buttons with nothing connected to them, bordered in all directions by webs of varying sizes. If you keep going, you’ll find that webs begin to become connected to other webs, resulting in larger and larger webs. You’ll also find that fewer and fewer islands remain. Eventually, the whole thing will be connected; it will become one big web. But it doesn’t happen steadily.
This random graph undergoes the equivalent of a phase transition when the ratio of buttons to threads reaches 0.5. So you can actually predict when the giant web will emerge! When there are 500 threads on the floor, something happens. Below 0.5, all you have is random assemblage of webs and islands – and the largest web is pretty small (a maximum of say 100 buttons). But as you approach 0.5, the webs get larger and begin to interconnect but there are still quite a few of them. But as the 0.5 mark is passed, whamo, the majority of the webs become interconnected, in one giant web. And it doesn’t matter how many buttons you use. If you throw 10,000 on the ground, as soon as there are 5000 threads, you can be sure that there will be a giant web. This is a classic example of a phase transition.
The key thing to take away from this is the idea that a phase transition can almost instantaneously change the face of things. Below 0.5, the random graph above is nothing more than a bunch of buttons and threads random connected and strewn about. Above 0.5, you quickly have a makeshift fishing net! Think about that. If you found this button and thread net hanging in your garage, would you think of it as a net or as a collection of buttons and threads? This is abstraction at its finest, and it happens via phase transitions. So what, right? Why should we believe they play a role in the origin of life explanations? Well let’s try a random graph with chemicals.
Considering how the random graph underwent a phase transition, let’s jump from talking about abstract non-living systems to talking about abstract living systems. A metabolic (or chemical) reaction graph is a graph with chemicals represented as circles, reactions represented as squares, lines indicating the reactions between chemicals, and arrows pointing to the products of chemical reactions. Here’s an example…
They come in handy when you want to visually represent all of the reactions that take place between a certain set of molecules. A reaction graph (or reaction network) is a good way to show what’s going on within a chemical system. For simplicity, we’ll look at an abstract reaction graph. Instead of using real chemicals and concerning ourselves with the specific details of reactions, this reaction graph will use generic chemicals and some simple kinds of reactions.
In chemistry, reactions are really nothing more than molecules breaking apart and mixing together to form new molecules or energy or both. Imagine chemical A and chemical B. You can turn chemical A into chemical B, and vice versa. These are one substrate, one product reactions. You can also combine A and B to get AB, and you can cleave AB to form A and B. Simple, you’re now an expert at chemistry. From there, you just add more chemicals and do more of the same kind of thing.
For example, look some more reactions you can get from As and Bs:
AB + A = AA + B
AB + A = ABA
AB + A = A + BA
I could go on and on showing the endless ways these chemicals could react to produce new chemicals. But that is not my purpose. The relevance of the reaction graph is that it works a lot like the button and thread network. Basically, you can think of the buttons as chemicals and the reactions as threads. Here’s what really matters: A network of chemicals can emerge from a random soup of chemicals simply by tuning the ratio of chemicals to reactions.
Instead of throwing buttons on the floor, let’s throw a bunch of generic chemicals into a beaker. If you’re dumb enough to try this, please place your computer next to the beaker so that there is a high likelihood that it will be destroyed if something goes wrong. I’d hate to get sued. Anyhow, for any set of chemicals, a chemist could predict what reactions would take place. The laws of chemistry dictate what the reaction graph will look like. Not too exciting really. But what happens when you add more chemicals to the mixture? You get a whole bunch more reactions. Things start picking up a bit. Thinking in terms of As and Bs, this makes sense. It’s easy to grasp that mixing AA and BB will have more possible reactions than mixing A and B. There are only three possible reactions when mixing A and B:
A can become B
B can become A
A and B can become AB
That’s about it. But look at AA and BB.
AA and BB can become A and ABB
AA and BB can become AAB and B
AA and BB can become AB and AB
AA and BB can become A and AB and B
AA and BB can become AABB
AA and BB can become A and A and B and B
It turns out that as the molecules get bigger, as you add Cs, Ds, and so on, the number of possible reactions increases exponentially. This is not only because of the reactions that can proceed between the initial chemicals. The products of those reactions then become the substrates for new reactions, thereby further increasing the number of reactions. So as you add more and more molecules, the number of reactions that can take place goes up very quickly. And just like the button and thread web, when the ratio of molecules to reactions gets to a certain point, the whole thing becomes an interconnected network.
Now for the big question: can we imagine a reaction graph for a set of chemicals that could lead to the origin of life? In other words, what would the reaction graph of the primordial soup look like? Since there is really no way of knowing, the best we can do is to explore the idea in generic terms to see if anything interesting happens.
Self-organization theory, by putting together the notion of phase transitions together with the complexity of chemical networks, actually shows how the interconnected network can become collectively autocatalytic (self-perpetuating, stable, and catalytically closed), meaning the whole thing can provide for itself, withstand being perturbed, and just keep on running – like a living organism. The secret ingredient is a decent helping of catalysts.
Without catalysts, the fact is that the network is pretty boring. Yes, once the ratio of reactions to chemicals gets high enough, the whole thing becomes connected. But just being connected isn’t enough to produce life. Chemicals just sitting in a beaker together don’t always react very quickly – even if they are connected. Like a bunch of shy kids at a school dance, they take a while to warm up to each other and start interacting. If the kids are too shy, the dance never takes off and everyone ends up going home early. Similarly, a beaker with a set of chemicals that don’t react very much isn’t going to lead to the emergence of life – that’s for sure. Thankfully, catalysts are big stars in the world of chemistry.
Catalysts push chemical reactions along. In the school dance, this would be the equivalent of a teacher convincing little Jimmy to ask Mary Sue to dance. So the reaction graph we’re after has to have catalysts. But since everything is connected, all chemicals are either substrates or products. That means that some chemicals will have to act as catalysts in addition to their day jobs as substrates and products. This is acceptable. There are plenty of examples of this in nature.
But the mere presence of catalysts still doesn’t get us to catalytic closure. In order for the system to become autocatalytic, a set of connected catalyzed chemicals must be present. Within the larger connected web of reacting chemicals, there must exist a subweb of catalyzing reactions. Catalytic closure, which Kauffman asserts should be a major component of any definition of life, means the system is continuously reacting using chemicals it has or produces. Little pockets of catalyzed reactions in the system won’t achieve this. The catalyzed reactions must all be connected to get closure. Luckily, with the help of our old friend statistics, this is not too hard to imagine.
The question now is what is the likelihood that a system with a connected catalyzed reaction subgraph would arise naturally in the primordial soup? Is it a fairytale or could it actually happen? To find out, we really need to know which chemicals can serve as catalysts in any given system. But rather than get tangled in analyzing each chemical, let’s just assume that each chemical has a one in a million chance of catalyzing any given reaction. As remote as these chances may seem, we can still easily show how increasing molecular diversity will inevitably result in the emergence of a collectively autocatalytic set.
Think back to the fact that chemical reactions increase exponentially as the number of molecules in a system increases. If you keep raising the diversity of molecules in the beaker, eventually the ratio of reactions to molecules will reach a million to one. Therefore, the average chemical in the system will undergo a million different reactions. So, probability tells us that each chemical will then catalyze at least one reaction (remember it has a one in a million chance). That means that the ratio of catalyzed reactions to molecules in the system would then be 1.0. At that point, it is highly likely that a large web will emerge, containing a fully connected catalyzed reaction subgraph. At that point, it will be collectively autocatalytic – and alive.
This explanation may seem too generic to be real but that is the point. The key to this line of reasoning is the idea that once any chemical mixture gets to a certain level of complexity, it is easy to see how living order can emerge. It needn’t necessarily even be organic; that just happens to be what was around on earth way back when. If we change the likelihood of catalysis to one in two million, well then the system just needs to have more living diversity, which is really just more time. The message is that the primordial soup had eons of time to work with. It is entirely plausible (and actually very probable) that the molecular diversity became sufficient to cause phase transitions that resulted in collectively autocatalytic, living systems. If this seems like the ultimate just-so scientific explanation, I understand. I’ve only scratched the surface on it. Stu Kauffman is the father of self-organization theory, so you can expect much better from him. Read his book for the juicy details – there’s depth to be absorbed.
My only aim in this lengthy discussion has been to propose a VERY plausible way to get around the supposed intelligent design problem. Once again, we are confronted with the limits of man’s imagination, not the limits of nature. So…I say again, please bring me an argument against evolution that holds water. Please.