The Origins and Nature of Life | Interview with Eric Smith | Biomusings

It’s complicated, right? you know the idea of low-hanging fruit, the easy problems that are the first things where you get a hold on. A lot of that is what has been done –for formal work– a lot of that’s what’s been done in physics. So, problems that have few degrees of freedom lots of symmetry lots of redundancy. And there are easy problems also in the living world. A lot of that was done in the original natural history: certain aspects of taxonomy, lots of the big phenomenology of living things.

But, if you ask: where do you get 01:13 - to a point where informal methods limit your ability to understand and you need formal methods to help? In physics that happens fairly early and so the formal methods get you a very long way, beyond what you could do with informal observation. In biology, we still are working so heavily in the characterization and the observation, that it’s much less common that we find formal methods are needed to show us what the concepts are. So, there’s a lot of quantitative work in biology but much of it’s modeling. And it’s a little bit different in physics, in physics the modeling problems were part of the job, they were what was often called phenomenology: the ability to express precisely what a phenomenon consists of and what you need to put in a model to be able to reproduce it. But, it didn’t require an enormous amount of phenomenology before the calculations started to show us that we needed new concepts that we had not seen before.

In biology there’s a huge amount of work 02:17 - just to be done on the phenomenology and the modeling, and it’s much more rare I think that that kind of formal work has shown us that we need a concept that we had no understanding of from the natural history side or from the informal side. Schrödinger’s “What his life?” might be the only theory book that was ever important for thinking about the origin of life. And the thing that’s fascinating about it is that Schrödinger had an amazing ability to see important questions, but almost every answer he proposed to those questions turns out to have been wrong in the specifics. So, the most famous of all the answers was the idea that it must be an aperiodic crystal that was carrying the information about life. And nowadays everyone looks at DNA and says “aha! Schrödinger understood what DNA would be”.

And even Francis Crick credited Schrödinger with part of the motivation 03:22 - to have looked for that structure in the nucleic acid polymers, but what people outside physics don’t know is that there’s a theorem, that in equilibrium systems there are no one-dimensional crystals. So the idea of an aperiodic crystal that’s one-dimensional, is already known to be a mathematical impossibility. The only way an aperiodic system can be part of a memory process is if it’s embedded in non-equilibrium, so that the theorems of equilibrium are no longer binding. And it turns out that if you ask: what do we know mathematically about how to make a one-dimensional memory system out of equilibrium? That’s an incredibly hard problem. If we want to understand what class of systems we have to look in for a molecule like DNA to fill the functions that it fills, they have to be a class of systems that are qualitatively entirely different from equilibrium systems.

So, in 04:29 - one sense that’s not necessarily a new physics, it’s an exten– it doesn’t overturn any of the physics of equilibrium that we know– but it shows that the equilibrium physics has very limited scope and there’s much much much more that needs to be added to it and it needs to be discovered. It’s hard to know how much Schrödinger could see the need for that, from a 1930s or 1940s perspective, but now looking back we can say “if he could have foreseen it, this would have been a good thing to be referring to”. And I think that’s often the way it is, you know, it’s like horoscopes: somebody who’s good at making general statements can say something so that no matter what happens you can look back and say “boy, they really had vision!” Some of it is knowing the right level at which to generalize and when to stop committing to more detail. So this… this requires care, because we inherit a huge amount of baggage in our science from a pre-scientific age, right? In a pre-scientific age and still today there’s a good contrast between alive and dead, right? So, if I see a seal on the beach I can say either that’s a live seal or that’s a dead seal. And, in either case, I start with the object and then I can put them into classes: these are the ones that are alive and these are the ones that are dead.

06:05 - We inherit some of that balance when we talk about things that are alive and things that are not in the category of ever living. And I think we try to imprint the alive and dead on the alive and never alive. And that’s where I think we make our first mistake. And you can see this in some sense with Aristotle, I don’t know this branch of philosophy well, but my sense is that he was struggling with what it is that makes the living systems all somehow like each other and different from everything else, but that doesn’t mean everything else is all the same; everything else has subcategories too. But –and I am getting to your question– what I think we inherit from that is that in all of those approaches we try to use life as a predicate.

06:56 - So, a predicate is a classifier, where we start with a set of entities and then we try to partition them to those that are in the class and those that are out of the class. And if we think we start with entities, then we try to partition them into the live and the dead or the living and the never living. We knew we were starting to get in trouble with that as soon as we started to understand viruses, because people said: “Okay, here’s the entity, is it alive or is it dead?” and those conversations never resolve. But then you think about our status in ecological systems; we have all sorts of dependencies on the ecosystem, and without it we have no status as living things, you put us in a vacuum or on a sterilized Rock and we don’t last hours, right? So, one of the ways people tried to get around that was to say “instead of a predicate, can we have properties? the living things replicate or the living things evolve” Well, that runs into a different set of snags. Which are, in order to talk about evolution you first have to be able to talk about individuals and populations, but they’re good arguments that there are important patterns in the biosphere that are fundamental to life, where individuality and population dynamics –even though they exist– are not the right level of structure to explain why the patterns are this, instead of something else or instead of nothing at all.

And so, all of these approaches to trying 08:26 - to define life run into one or another problem, and I think the problems are fundamental, they’re not merely technical. So, this is where I think the collaboration between the sciences can do something that’s new and better. So… in my opinion the most important accomplishment of 20th century physics was that it taught us how to think about entities and interactions inherently in statistical terms. Before then we had always thought in terms of objects and some other abstraction for motive, but there were always problems with doing that, because things we thought were objects were not ultimately objects –they were composites, they were something else– and we had no language in which the central tendency of a fluctuating thing could fill the role that objects had always filled. The reason I say this is the most important thing in 20th century physics –more important than relativity or quantum mechanics– is that relativity and quantum mechanics were both insights to how a particular part of nature is built.

09:34 - But they applied to that part of nature, even big as it is. The insight about how to understand that statistical concepts can provide a language of objects and interactions is true for all of those phenomena and also others that the physicists have not been looking at yet and that we study in the other sciences. So, if we say what can we learn for that when we come back to biology? Can we address the nature of life in a different way? You can say, okay, if you want to understand a system you want to understand all of its essential properties. In the living world, some of those are things that are shared by all living things they don’t change and others are things that change all the time and it’s the ability to change that makes life interesting. Evolutionary dynamics has focused very extensively on the things that change and that have the capacity to change, because that’s what we can say something interesting about.

10:32 - They have given far too little attention to the things that never change like certain patterns in core biochemistry, certain aspects of cellular structure and cell physiology. And if you look at those, the things that look more law-like and less flexible, you start to find deep connections to geochemistry and things that don’t look like they’re best explained by the phenomenon in population biology. And if you say, okay it’s not like this is life and that is not life –they’re all parts of life–, what is an abstraction of life that somehow captures all of this? What I think the right thing to do is to distinguish the trajectories of planets. To say, if this planet had no biosphere what would happen and what would never happen? The planet that we know has a biosphere and there was a sequence of transitions through which the biosphere came to be the structure it is today. What does the biosphere do and what does the planet do –including the biosphere– that could only happen with the structures of the biosphere? And then, can we describe that in statistical terms? The same –broadly speaking, mathematically– the same way we describe the difference between liquid water or a unmagnetized piece of iron and frozen crystallized water or a magnetized piece of iron.

A kind of pattern comes into existence 11:59 - that did not exist before. In the biosphere the patterns are dynamical and they’re chemical and they involve hierarchical structure and they involve self-perpetuating patterns. What we want to understand is that the statistical central tendency for a planet with no life lacks all of a certain class of patterns, and the central tendency of a planet with a biosphere suddenly has those patterns as the durable robust things. And everything else all of the structures whether they’re cells, or viruses, or ecological interactions, they are part of the architecture that participates in that dynamical maintenance. But you don’t want to mistake the architectural feature with the statistical central tendency which is the durable thing.

12:46 - So, it sounds like a very roundabout answer, but the answer I would give you is that the nature of the living state is not a property of organisms the way being alive or dead is. The nature of the living state is carried by the biosphere as a whole. And then we ask, why is organization into genomes and cells and organisms and virions the robust architecture? A lot of times when people think about origin of life they mostly think about wanting the historical story: the sequence of events and places and changes and so forth. And it’s important, but to me the thing that is more interesting, even than that, is where the exercise of reconstructing the history tells us that we need to rethink the concept of what life is. And this is where I think the the object language that’s traditional in all sciences is the thing that we most want to get beyond in biology.

So, in biology for a very long time, 14:01 - organisms were kind of the fundamental objects. And if you read Stephen Gould’s structure of evolutionary theory, he argues in the introduction that Darwin had a huge commitment to object –to individual organisms– as the central points of change for living order. And even though he was a consummate ecologist, he understood the essential character of relations, he felt that the science that he could make was a science of selection of individuals. As we learned more about individuals than was known in Darwin’s time, we realized that they’re divisible and that they’re changeable and they’re assembled and so forth, so they weren’t workable as fundamental objects. You have this classic book by EB Wilson “The cell theory of life” –I don’t know if that’s the title or not– but he makes the argument that the cell is the fundamental architecture of all of life and that that architecture is more fundamental to the nature of the living state than any of the properties that particular cells have.

But we keep chasing the object 15:03 - language down and down and down one level and the result, I think, is that it’s led us to a skewed biology, where objects are at the center of biological theory and other domains like ecology are kind of orbiting around them in a loose way –their community assembly or their some science of relations– but if you do biochemistry in deep time, you start to see features that don’t, they’re not as essential properties of individuals as they are properties of ecosystems and you say, in biology we need a role or a theory of ecosystems as fundamental entities of their own kind, because whatever those entities are they seem to be the things that are carrying the univerisality of metabolism, and it’s connection to geochemistry in each era, and the major transitions in the forms of biochemistry through time. And then cell form can change, enzymes can change, genes can change, the composition of chromosomes and species can constantly turnover. And yet these invariants always are kind of the reference point that they support at the ecosystem level. And this is the thing that has pushed me to say: we need to move away from the object language, we need to start to understand how patterns in biology can have an entity status even when they’re they consist of dynamically preserved relations and there’s not an object per se carrying that relation. That’s a huge departure from anything in physics, because in physics if you ask, where is there a durable pattern like the crystalline form of a diamond or zircon or something like that, those patterns persist because the objects that carry them are durable objects.

16:53 - In biology and biochemistry the objects keep turning over but the patterns keep being created. So what we need from this kind of new collaboration of physical insights and biological and chemical insights is a theoretical language that enables those self-perpetuating patterns to do what entities did for us. The great confound in biology is between things that are universal because they’re the only solution to some problem and things that are universal because they’re inherited from a single ancestry of life. And typically people go too far off one side or the other, like they’ll see everything Universal and say it probably is all the only solution that’s possible, or they’ll see everything that’s Universal and say it’s all a frozen accident and it could have been completely different. And Francis Crick and Steve Gould are probably the best-known people arguing for the importance of accidents in biology, because they were trying to get away from a kind of Laplacian view of the world that made too strong an assumption that everything was governed by laws that only had one solution.

And they 18:08 - fight hard, because they were trying to get a new point of view into people’s consciousness. But as a result they take a position that in my view is too extreme, so, if you have a thing that is the way it is for a law like reason and you think it’s an accident, then what you’ve done is fail to even try to understand why that was the only solution. So, I think the right middle ground is to say, the elements in the periodic table are really all unique and the chemical things that each element can do are not exactly duplicated in any other element. For some of them the changes are huge, like carbon- nitrogen -oxygen have characteristically irreplaceable roles in biochemistry. The relation of sulfur to oxygen, the relation of phosphorous to oxygen and to nitrogen dictates much of their roles.

You go down 19:07 - into the metals they start to be a little bit more interchangeable –they as a group have roles that nothing else can substitute for– but individually you start to see some substitutability. So that’s law-like because quantum mechanics doesn’t give you any choices. At the other end, there are five fingers on my hand and we know that my deep ancestors had different numbers of fingers, so we know that that’s not governed by a law. Somewhere in between, where does law hand off to things that get locked in? That’s probably not a simple boundary, it’s probably a kind of a ragged coastline. And so, for me, the thing that makes biochemistry so fascinating is that it is that mix of things that probably only have one solution.

And, that solution keeps being 19:52 - supported by higher-level machinery of different forms, but it’s always the same solution. And, you know, if you look in deep biochemistry, the biggest candidate for that is the citric acid cycle and the role of the simple amino acids, there are no substitutions for what they do, in all of life. Everything… all other phylogenetic relations can change, those roles are invariant. But then you get up into the more complicated amino acids and some of the, you know, some of the sugar chemistry or the hydrocarbon chemistry. Some of it looks like it’s anchored to things that don’t change, other of it looks like it probably comes from accidents of how catalysts were available or things like that.

Can we understand where that boundary in that trade-off takes place? 20:36 - To my mind that’s what we could do to get a full and balanced understanding of the different things that contribute to the living state. There’s an important thing to say as part of building a better scientific dialogue, which is the way that reductionism is used in biological sciences –which they would like to blame on physics– is completely different from the way reductionism is used in physics at its best. It may have been the way it was used in the physics of a hundred years ago, but that’s a thing that I wish we could get beyond. Because the major accomplishment of the 20th century in physics was the understanding of the nature of emergent order. And we now realize that reductionism never made scientific sense without an understanding of emergence.

Because, if you didn’t understand how a new kind of 21:39 - order can shield the details of smaller things, reductionism would lead you to need to know everything to predict anything and the point of emergence is that you can have a predictive finite science at different levels, which lets through some of the details from the foundation and masks others. And the important thing is, at the boundary where that happens, you have systematic ways of learning which details will come through and which will be masked. And a good reductionism in every science I think is going to have to have that dual character. The essence of reductionism is that, if a composite thing is made of building blocks, then the number of combinations will be much larger than the number of kinds of building blocks. If all you want to know is the properties of the building blocks then the number of experiments you have to do to get to answer that question should only be about as numerous as the building blocks; it should not have to be as numerous as all the combinations.

22:43 - And that’s been incredibly useful in physics because it says we don’t need to know everything about chemistry to work out the periodic table, and accomplishing and understanding of the periodic table is finite progress, even though we still don’t know most of what’s in chemistry. But you would never say that knowing the periodic table we now know everything in chemistry, that would be stupid, because it’s manifestly false. So, to understand the combinations you need a new native language that is addressed to the characters of the combinations. And the knowledge of the building blocks does not give you that language, because the combinations are a new kind of intellectual entity. So, in biology, the role of reductionism is to know when you can solve part of a problem without having to solve the whole problem of the combinations.

But in 23:30 - no mature science would you ever think that that means you don’t have all the rest of that work to do, so we shouldn’t think it in biology either. There are certain kinds of pattern that you develop a skill for seeing from the kinds of systems that physicists have studied. And there are different kinds of patterns and combinatorial spaces that you get good at searching and you develop a feel for if you spent your life becoming a chemist and then there’s yet another class of patterns that you become good at feeling your way to if you’re an evolutionary biologist. And it’s just like learning a second or a third or a fourth language there are developmental periods when you you become really good at something that you spend a lot of time in and if you go into these other fields afterward you miss that window and there are things that the experts in that field can see that you never become as good as they are at seeing and so when we work together we just have to understand that every one of us carries these vast blind spots and we got a we got to trust that the other ones see how to do things that we cannot see and we may never be able to see. But this is where it gets interesting, we have, you know, ideas come from human minds, they come from the flash of recognition within one mind and yet carefulness comes from covering a whole field.

We have to do carefulness at the level of being an expert, but we have to 25:05 - cover so much material no one can be an expert. How do we build a community that gives the combination of those things? That’s a new problem. I didn’t say anything about chemistry, but if you had to say what’s the other really important thing about the nature of life that I think doesn’t get enough weight in most of the conversations, there’s a kind of a split in the way people talk about chemistry. A view that is sometimes criticized is people who, is an old version of saying “life is a chemical phenomenon”, and often the people who said that sort of meant we need to understand every chemical detail or we want to understand life. And there was a kind of a backlash against that saying we can’t understand every chemical detail, so many of them change.

Every time you say something is universal in biology, some biologist 26:02 - will find a violation, because that’s what they do. There are actual things that biologists haven’t found violations to but they don’t like to admit that there are. So, you don’t want to get hung up on every detail when the theme of biology is that so many details can change. But the backlash went too far the other way, where there were people who wanted to say “life is a kind of pattern and it’s only accidental that it takes place on a chemical substrate or that it takes place on this chemical substrate”. And I think that’s another kind of a mistake because there are certain things that the chemical substrate does that no other level of physical organization does.

And I think, you know, we now know 26:47 - what? 27 orders of magnitude in physical energies? We know a lot of the structures that physical systems can form. And there is this unique structure in the middle, that’s called chemistry, that has a capacity for complexity and for separation of time scales, for problems in integer arithmetic, and in group theory, that do not occur again anywhere else in the hierarchy of physics. And I can list a long set of things in the nature of life, it’s error correcting capacity, the structure of metabolism, where I think if you didn’t have chemistry to build from, nothing else about physical dynamics would ever make those patterns possible. And the particular forms that they have are tied to very particular places they occur in chemistry –now I can’t prove that, that’s a prejudice that I have that I want to turn into a scientific claim–, so you know you can take it for what it’s worth. There’s a kind of internal complaint that America somehow has the noisiest school buses in the entire world, and I don’t know why.

In any other country we 28:14 - would not subject children to this. .