There's been a huge variety of ideas throughout history about our importance in the universe, and that's one reason why the topic of aliens is important. Also it's fun, and also we want to know what we might meet some day… although we might also be interested in all the aliens that are too far away to ever meet. When we get around to estimating the life in the universe, I'm pretty sure we're going to decide that there's lots, but it's not so clear whether there's lots within easy reach.

Chapters 1 & 2 examine the basics of life on earth.

We have chosen to define some terms here. Homobiota is earthlike life, which uses extremely similar biochemistry to our own. It is carbon based in an aqueous medium, it forms itself from DNA, RNA, lipids and polypeptides, and most life utilises membrane compartments such as cells. Heterobiota is life that is not homobiota. Earthlife specifically refers to the life found on earth. We did not rigourously define it in terms of whether it was homobiota or a shadow biosphere of heterobiota.


This section contrasts the position that life is fundamentally earthlike, requiring carbon and an aqueous environment, or even requiring specific earthlike biochemistry, with more permissive perspectives that hold that the specific biochemistry of life on earth is emergent by the constraints of our biochemical environment and history.

This changes what we look for when we look at places other than earth for signs of life.

There are modern discoveries of biochemical unity of organisms on earth.
"The complete set of chemicals utilised by Earthlife, however, represent only a very small fraction of the vast number that may be produced by natural processes." - Links to the possibility of other Life out in the universe, which could be completely different chemical to what we're used to.

There is now experimental evidence supporting some of Feinberg and Shapiro's arguments that our particular biochemistry is not special, e.g.: (I haven't vetted this yet but on a cursory reading it seems like a reasonable summary)
Modeling paper:


F&S review proteins, sugars, lipid membranes, DNA, RNA, NTPs, and the machinery of transcription and translation
They not the arbitrary stereochemistry (right-handedness) of biochemical molecules

=> The arbitrariness of stereochemistry is an example of the super-general point that evolutionary processes can fix an arbitrary decision so that it becomes the basis of a huge amount of biology.

Other examples:

=> those in favour of aliens NOT having the Earthlife characterisitics (e.g. cell walls) is becoming more agreed upon as we discover more science

F&S note that things needed in one part of the cell are produced by another part, a complex cycle, and they briefly explore the structure of atoms.



"Energy" is actually a metaphor for something really complicated and I'm not even sure what, unless maybe it's entropy. Because the whole physical world has plenty of energy. The problem that biological "energy" mechanisms solve is making it available for specific types of things like endothermic chemical reactions.
All (pretty much all) energy used in our biosphere originally comes from the Sun. Earthlife needs a constant supply of sunlight as the energy already on Earth is converted to thermal energy, which is very difficult to use.

At the risk of sounding very hippy, the biosphere itself is a living organism with cycles of its own. Organisms on Earth are cells in a superorganism. I don't remember if there is a specific term for it, but I like this framing; the notion that there is a "geostasis" maintaining the biosphere in the same way that, say, homeostasis maintains body temperature). There will be more about this later in "Life Beyond Earth".

Living things on Earth constantly take in useable energy (some of which is necessary for living and some of which is expelled as heat) - although this really means entropy or something similar - living things elsewhere in the universe would have to do this too - second law of thermodynamics

Examples of Energy on Earth

ATP is a universal energy store in Earthlife. This makes all life of Earth incredibly similar, as ATP could easily be replaced by GTP, CTP, TTP, UTP, or even other imaginary energy sources. We can easily imagine squillions of alternatives. The alternatives might not be as good as ATP ... but then again (a) they might, and (b) we shouldn't be so sure that ATP is wonderful, because evolution doesn't generally come up with optimum solutions to problems. We'll talk more about this when we do "The Extended Phenotype".

What is life made of?

Earthlife is made of arrangements of atoms. Atoms are arranged through bonds, with arrangements differing depending on the specific connections between atoms.

Atoms are exchanged through cycles of elements (pg. 74)
e.g. carbon found in leaf -> eaten by insect -> insect eaten by bird -> animal -> animal -> sea -> limestone -> erosion -> ...(etc)

Possible language: Living things would need to sit along some kind of gradient from which they can harvest energy. (I like this. TK Off topic, for Ada: since you think that the arrow of radiation is different from the arrow of thermodynamics, you might have an opinion about whether life needs an entropy gradient or whether a radiative gradient would do.)

CHAPTER 4 : How Earthlife Evolved

Formation of the Earth and Emergence of Life

Big Bang (13.8 billion years ago)
Life as we know it is very recent (less than 4.5 billion years) in comparison to the history of the universe (eg. plants, animals came into existence much later than the big bang)

Early universe not compatible with life made of modern matter, and similarly life from the early universe (eg: quark-gluon plasma life) would likely not survive in our modern big cold universe.

According to F&S, the Earth forms ~ 4.5Gya, and The Earth's crust solidifies ~4Gya. I think these numbers have changed since then but not in a way that is conceptually different..

F&S use an analogy of the history of the Earth scaled down to the course of a week, similar to the way they created COSMEL for grappling with big numbers and issues of spatial scale.

Chemicals required for Earthlife could be formed spontaneously through a number of processes. In the Miller-Urey experiment, molecules of 10-20 atoms were created from sunlight passing through the gasses that were available on earth like methane, carbon dioxide, water vapour, hydrogen and otheren. Some of these are used in Earthlife. It is possible that other planets with similar conditions would have made use of different molecules that Earthlife, so even worlds with similar conditions could have radically different life.

To work out how old life is we rely on fossil remains - but there are some issues with this. We can also do phylogenetic analysis, but this has some assumptions and limitations that cause more and more uncertainty the further back you look.

F&S disagree with Dawkins's statement in The Selfish Gene that life starts with the replicators. They seem to be arguing that while replicators are important, life starts before them (p97). So for sure we'll come back to this when we do The Extended Phenotype (which recaps and overlaps with The Selfish Gene).

The closest relative of the Eurasian mole is a wolf, and the closest relative of the African mole is an elephant. Or something equally amazing. TKTK Ada wants to look this up and make sure this statement is actually true. But if it isn't a literal truth it still describes a true concept.

Most of Earthlife is single-celled (actually, on consideration this is probably not true if you count viruses as life; I suspect virions outnumber cellular life TK), and for most of history (3/4) all Earthlife was at most single celled. I believe Stephen Jay Gould used to say that "Bacteria are the real lords of creation".

When we think about the proportion of life, we get this wrong and tend to think that big multicellular life is more common. Stephen Jay Gould argues that this overweighting of complexity is because we don't realise we're only looking at the tail end of the bell curve. As complex things become more complex, the tail end gets stretched out towards more and more complex life, but the median doesn't really move.

Oxygenic photosynthesis

At some point, microorganisms would have depleted their stores of chemical energy. Photosynthesis allowed them to get around this by converting solar energy into chemical energy. Oxygenic photosynthesis, in particular, makes a large amount of energy available because it works through oxidation of water, and there is REALLY a lot of water around. This caused a change to the biosphere, first a buildup of oxygen, and then the formation of an ozone layer that protects life from UV radiation.

We expect that most forms of life around at the time would've died (this is the great oxygenation event, which was a mass extinction), isolated themselves from oxygen by burrowing into the plegaic mud, or adapted to survive with oxygen.

The abundance of oxygen also leads to a new chemical process to obtain energy: respiration. The energy stored in the chemical bonds of molecules can be harnessed through a stepwise chemical process, making a great deal of energy availible.

F&S argue that predation becomes possible with the advent of photosynthesis. The argument goes like this (p99):

4.1.1. Oxygenic photosynthesis makes free oxygen availible.
4.1.2. The existence of free oxygen makes respiration a viable reaction to use to harvest energy.
4.1.3. Respiration makes catabolising the constitutive molecules of another organism a viable spource of energy.

This is reasonable, but I have issues with it. Even prior to respiration, surely there would be evolutionary value is hunting other organisms for parts. Perhaps you can't trivially harvest energy by catabolising their parts, but you still save yourself a bunch of energy by stealing their small molecules rather than synthesising your own. In theory one still needs an energy source for one's own processes, but if you're stealing the building blocks from other creatures you could also steal their energy stores. ON THE OTHER HAND, are there any examples of actually predatory microbes preying on other microbes in modern Earthlife? It is possible the maths just doesn't work.

As an extension to this: presumably pre-respiratory organisms were able to store energy in some way, such as as ATP, and I think they would have developed this very early as one needs a supply of energy inside a cell to power biochemical processes. (Do you actually need this? I think it is important, you harvest your external entropy gradient to build an internal entropy gradient that in turn powers your biochemistry, rather than directly linking your biochemistry to your external entropy gradient. This is important for biostasis; it's unhelpful if your organism dies because your energy harvesting process briefly stalled halfway through replication and you couldn't finish your DNA. Furthermore, most of the biochemical reactions are driven by this internal energy gradient, so you would need to couple EVERY biochemical process with your energy harvesting process. This seems unfeasible; you can't have every biosynthetic enzyme directly linked to a copy of your energy harvesting enzyme, you'd need far too much harvester.)

F&S talk about oxygenic photosynthesis allowing colonisation of the upper ocean and the land. The argument goes like this (p99 to 100)

4.2.1. UV radiation is quite high energy, and causes chemical changes that disrupt the biochemistry of living organisms.
4.2.2. Oxygenic photosynthesis leads to a buildup of free ozygen in the atmosphere.
4.2.3. Free oxygen in the atmosphere is converted to ozone through some process (I assume UV catalysis)
4.2.4. A layer of ozone soaks up high energy UV rays, preventing them from reaching the surface of the Earth.
4.2.5. In the absence of the destructive high energy UV rays, it is now easier to exist on the surface of the ocean and even on land.

This hypothesis is well regarded, and I accept the idea that the creation of an ozone layer made the colonisation of the upper ocean and then the land much easier, leading to an explosion of life in those spaces. But the framing is that it made this colonization possible and I don't know that I agree. My argument goes like this:

4.2.6. By F&S's count, there are two billion years from the existence of bacteria and algae(?) to the formation of the ozone layer. That is an exceptionally long time on an evolutionary scale.
4.2.7. The danger of high energy UV provides a selective pressure for an organism to be able to deal with the damage due to UV radiation.
4.2.8. The energy availible in high energy UV provides a selective pressure through the opportunity for an organism to be able to utilise that energy.

We know that there exist biological processes that clean up DNA damage due to UV radiation (yay \\deinococcus radiodurans \\ ), and that some of these processes are even powered by energy harvested from UV radiation. It seems unreasonable to assume that life prior to the ozone layer would not have been able to develop similar processes.

We know that dangerous selectionary pressures (eg, the radioactive center of chernobyl\\) provide a strong selective advantage to organisms that can survive them, as if you are the only thing that can live in a particular niche you have much less competition.

Given this, why do we assume that there was no life in the upper ocean and on the surface of the Earth prior to the ozone layer?

Is it that the intensity of cosmic radiation was such that it would not be possible to circumvent or harvest (I haven't done the math on this, and I know the energy of cosmic UV bombardment is A LOT and destructive, so I'm willing to consider this view)?

Is it simply that we have no evidence that these things existed? (I suppose there may be geological evidence that nothing was eating oxidisable minerals on the surface of the pre ozone Earth), but also it is plausible that whatever adaptations were necessary to survive in the presence of high UV energy might have lead to an organism with a smaller paleological footprint; I confess I don't really know anything about paleology of microorganisms. \\ Paleology of microorganisms is a really good topic. I know that most of our paleology is from bones + shells - dinosaurs and trilobites, basically; I also know that we have some from soft tissue including bacteria. A possible topic is for someone to look up more about this.

\\ It would probably be prudent for me to look up whether anyone has actually run math on this, because surely they have.

The UV high surface of ocean area is very difficult for life with DNA to live in, so there was possibly very little competition in this area. In two billion years life that is not DNA dependant could have evolved and colonised this area free from competition and then died when the UV levels went down without leaving a trace. However, there is a question of scale. If we assume that life starts as some weird chemical hodgepodge and eventually reifies into a consistent thing (eg: RNA), that will take a certain amount of time. We have a problem tho, in that the weird chemical hodgepodge might be consistently outcompeted / devoured by DNA life, so once consistent life has evolved once, it doesn't get a chance to evolve again. (Eg: endosymbiosis only happened once for mitochondria in eukaryotes for this reason; other endosymbiants were outcompeted by the descendants of the first one. Ball and Brindley talk about this a bit.

Evolution can be thought of as burning a succession of small bridges: the results of a transformational evolutionary step usually destroy the preconditions for its own occurrence.

This might not be a problem if our new life in the upper ocean is isolated, but I don't know that we can assume that, it might be swept into and out of the extant life's biosphere e.g. by ocean currents, and so it can't remain isolated. And in general, life-not-related-to-us is likely to have evolved (given the immense time available) provided it had a niche where it wasn't immediately outcompeted by life-like-us. We don't know for sure whether that ever happened (but I bet it did, at least for a little while).

Jason notes that this is like gut bacteria; even if 99.99% of our nascent surface life gets swept into the "be eaten" zone it's fine as long as the remainder survive, because of the lack of competition coming in from outside (just like relatively few new acid-resistant bacteria get added to our guts, or in the ocean case the lack of competition would be because UV quickly kills incomers). Ada has concerns about feasability w/r/t diffusion but can't quantify them right now. TKTK

It is possible that there was a bottleneck that prevented this evolution (cf. whales don't have gills); eg: it might require that you stop using DNA (big problem) or the energetic expense of UV protection might be prohibitive.

Here we discussed the idea of an adaptive landscape, and the notion that many processes (adaptation, protein folding) can be viewed as optimisation or energy minimisation problems.

CHAPTER 5 - the origin of Earthlife: An unsolved problem


5.1. Spontaneous generation- the idea that non organic things like mud will spontaneously become life in the form of animals or single celled organisms
5.1.1. Spontaneous generation of microbial life - very convincing doubt was cast on this :-) by Pasteur
5.1.2. Carbon compounds necessary for Earthlife are spontaneously generated through natural processes.
Carbonaceous chondrites are meteorites containing significant percentage of carbon in the form of organic compounds.
The murchison meteorite contained racemic mix of L- and D- amino acids. The mixture suggests these were not of Earthlife origin.
Spectroscopic (spectrographic?) analysis of radiation from dust clouds allows identification of compounds in the dust. At the time of printing, the largest organic compound identified had nine carbon atoms. Since then, we have found larger compounds, including C_60 and C_70 fullerenes

5.1.3. Prebiotic soup
Theory that the seas of primitive Earth accumulate organic compounds until they were a soupy mixture of many chemicals.
It's possible for organic compounds to build up because there are no living things to eat them and no free oxygen to oxidise them. Sources of these compounds could include (inter alia) UV radiation, heat from vulcanism and radioactive decay, shockwaves from thunderstorms and meteorites, lightning strikes.

5.2. Randomly made replicators
How does the first replicator arise from this mass of compounds? Can it just happen randomly? (One answer is Ball and Brindley's paper - see our entry about that.)

Problem: the soup wants to move towards equilibrium. (Consider though thermodynamic versus kinetic stability; some steps towards equilibrium have very high energy intermediates, and the move towards equilibrium happens very slowly. For example, polypeptide chains in water are thermodynamically less stable than their hydrolysed monomers, but the energy required for spontaneous hydrolysis at neutral conditions is large - ie: the polypeptide is kinetically stable. This is why our hair does not decompose when it gets wet.

The odds of replicator forming from random chance from a system at equilibrium are extreme; F&S estimate 1 in 10e-1000000. But obviously the initial system is not at equilibrium; energy added to the primordial soup peturbs equilibrium.

F&S raise the problem that the chemical space is vast, which problematises spontaneous generation of particular compounds like nucleic acids. (Presumably the favourability of some forms over others will address this up to a point, but you still need another selective process. EG: Ball & Brindley note the selectivity due to HP crucible.)

The origin of the first homobiota is very unlikely. All current specific models show that generation of the first life takes much longer than the Earth or even the universe has been around. This makes some "pessimistic" scientists found an argument for alien cells or for God. And they might be right, or it might be that we need a more specific model (like in Ball and Brindley's paper, but with even more details).

Spontaneous generation of the right chemicals for replication in sufficient concentration to replicate is statistically unsound.

Ada notes that this section assumes a purely random process, however these processes are often not random. Protein folding occurring randomly would take longer than the age of the universe; folding happens quickly because it's not random. Some states are more favourable than others, and some early states in turn make others more favourable; folding is an energy minimisation process.

5.3. Predestination
F&S discuss Predestinistm, a belief that all or at least some of the events leading to the development of current day Earthlife are inevitable given the fundamental laws of the universe.

In other cases, however, predestinist expectations can be compared with experimental fact. When this is done, they are generally found wanting.

They discuss the Miller-Urey experiment, which lead to the spontaneous generation of organic molecules. Here, they again define spontaneous generation as organic life molecules arising by random chance, later on they consider the emergent systems that tend towards something (ie: less random processes). They argue that spontaneous random generation of nucleotides, let alone polymerisation of them, are not likely to arise through purely random processes.

SOCIOLOGY OF SCIENCE: They note there is a problem in that negative results are not published, so we don't have a survey of mixtures of gases that do not give rise to spontaneous amino acids.

5.3.1 Cosmic predestination

Here they examine the argument that the fundamental constants of the universe bias the universe towards the existence of life.

They quote an argument on p.129:

The implication of these results is that organic syntheses in the universe have a direction that favors the production of amino acids, purines, pyrimidines and sugars: the building blocks of proteins and nucleic acids. Taken in conjunction with the cosmic abundances of the light elements, this suggests that life everywhere will be based not only on carbon chemistry, but on carbon chemistry similar to (although not necessarily identical with) our own.

They object to this on the basis that it's clearly nonsense.

They object to Hoyle and Wickramasinghe's suggestion that influenza could have arrived by panspermia; as our understanding of the mechanisms of viral infection grows more detailed the notion of a space born viable influenza strain virus seems even more unlikely. I like their quote:
Its likelihood ranks with the probability that Christopher Columbus, in his landing on San Salvador, would have found that Spanish was the indigenous language of the natives

Taking the fact that the constants of the universe allow homobiota to arise as having some significance or causal direction is a bit silly because the same argument applies to many things that exist; of course the constants of the cosmos perfectly match the conditions for homobiota, because homobiota is what exists here. If they were different, there would be a different homobiota making the same argument, or there would be nothing to make an argument at all.

If the Battle of Waterloo had had a different outcome, it is unlikely that any reader of this book would exist, as the particular combination of events that led to each of us being here is so unlikely that any large change in what happened in 1815 would almost certainly have led to a different group of people being alive today. But it would be absurd for reader John Jones of 12 Cedar Lane, Topeka, Kansas, to infer just from his existence that the Battle of Waterloo took place and was lost by Napoleon. It would be equally absurd to try to infer from the Battle of Waterloo that Mr. Jones would be reading this book 165 years later. Only after we know both of these facts are true can we construct a link between them.

‘The anthropic principle’ takes the stance that the universe is in a set on conditions primed for making intelligence in humans. This is taking a small view of what life could be (and maybe what intelligence is). This is a very popular idea these days.

5.4. Panspermia

Panspermia is the theory that life on Earth was seeded from a non-Earth source, such as by arriving in cosmic debris. It's pushes the origin of life backwards a bit, which is good if you think life couldn't have arisen in the available time.

Panspermia first proposed by Arrhenius in 1907. They also note theories where planets were intentionally seeded life (Directed Panspermia); with Crick and Orgel proposing it in 1973.

"Sagan has demonstrated that it is unlikely that even a single meteorite from another solar system has reached the Earth during its entire history." (p.134)
And yet a meteorite from another solar system was spotted a few weeks ago! It didn't get all that close to hitting the Earth, but still I think Sagan's estimate is probably wrong. My guess is that he used a massive underestimate of the proportion of stars that produce meteors.
(By the way, even if his point is right for the Earth, it would only show that panspermia doesn't happen much around here. In the centres of most galaxies, stars are packed much more closely, and so panspermia is much more likely. That's not relevant to what this chapter is about, though.)

5.5. Chemical evolution

Gradual chemical evolution says that some compounds are favoured by catalysts, and as they become more prominent, they are created even more-> is this the first type of replicator?

They seem to be arguing that the first type of replicator refers to the first thing that can replicate itself, and they are viewing a chemical catalytic soup that biases the local chemistry to create more of itself as being a different thing. Maybe because the soup is a lot of bits working together rather than one thing specifically copying itself? I don't know if I like that distinction tho, because a lot of things we might thing of as replicators can be viewed as a complicated mess of stuff rather than one distinct thing copying itself.

Oh I very much do not like that distinction, because very few organisms can replicate themselves. The Dawkins book is good on this (chapter 5, "The Active Germ-Line Replicator"), in case it's not already obvious. Also, this is why I don't like the standard reason for saying that viruses are not alive. ("They can't replicate except inside complex organisms that provide them with reproductive machinery." But nor can lots of things, e.g. many bacteria and even some wasps.) I suppose there is an argument to be made that this type of self organising system isn't replicating itself, rather it's changing the environment to be a thing that is more like itself. I'm not sure there is a substantive difference here; I want to think about it some more. Yes let's think about it, but I'll be amazed if there is a substantive difference. (Talking of substance, that argument sounds Aristotelian!) (Same)

5.5.1. F&S (although not other authors e.g. Ball and Brindley) think there was a period of simple, non-Darwinian evolution on Earth that produced a good mix of chemicals from which RNA could be produced ... and then Darwinian evolution could take over. This early chemical self-organisation on Earth was different to Darwinian evolution, F&S argue, because advantageous changes were not propagated by reproduction ... the point is that there is no process in this self-organising analogous to mutation and hence nothing analogous to Darwinian (open-ended) adaptation.

5.5.2. Different to predestination
They argue that this is different to predestination because predestinists hold that the universe is biased towards homobiota, or things very like homobiota, whereas this self-organisation would just give rise to some arbitrary selection of molecules, and there's no particular reason to assume that ours are special.

They argue that for such a self-organising system to exist you need a system far from equilibrium, and subjected to a flow of energy. They note their may be other requirements that are not obvious. I note Jason has previously made the point that we are better to think of the energy source in life as an entropy gradient rather than an energy gradient, I think that holds here too.

They note some early work on describing mathematically self organising systems suggetsed similarities, but the field seemed to be in its infancy. I wonder how that has progressed since the time, I'm trying to find a paper recently that suggested the emergence of replicators in some types of chaotic systems was inevitable. TKTK

Jason raises that the framing of predestinists as dogmatically focused on homobiota is not unreasonable as that particular view was quite commonly held (and it is still held by many) so F&S's framing could be reasonable (though it's hard to say whether they're right).

Some predestinists are def wrong because they're assuming a narrow definition of life. Others believe in a wider definition of life but think it won't happen ... although we have already artificially made other kinds of life!

Jason's initial thought was that we shouldn't believe predestinists because they've obviously got a suspect motivation (they want our kind of life to be special). But even if that's true (and it is true for some of them but maybe not for all of them), it's not nice to be using a psychological argument to say your opponents must be wrong, without looking at what they say. So Jason sucks. So far. But here's another argument for thinking the predestinists are wrong. It's an inductive argument that's like a Zeno's paradox in reverse:

We are getting closer and closer to discovering or creating the types of life that predestinists have said are impossible. The closer we get, the more likely we are to beleive that these types on life are actually possible but we just haven't seen them yet. And so we should stop believing something right before they concluded that the next thing is impossible, because is may not be (we may just not have found it yet) saying that a unicorn could exist is not the same as saying that is does and that we have to go and find it. I mean, narwhals exist and horses exist.

So that should satisfy us that heterobiota are possible, and when we do the numbers later (under Drake's Paradox) we'll see that unless they're super incredibly unlikely (which they're not) the universe must contain lots of them.

It's like if someone says there are no platypuses in the ACT, and you've never seen a platypus in the ACT but you keep seeing platypuses in NSW that are closer and closer and closer to the ACT border. You should end up believing that there are platyuses in the ACT, even if you never see one. This is a good analogy, because we keep (a) finding and (b) creating life that's more and more different from what we normally think of as typical life and closer and closer to breaking standard definitions of life.

A similar argument could be made about the existence of planets, as we find a much larger range of planet sizes and arrangements that we initially thought likely (eg: lots more hot jupiters very close to their star than we expected)

Even if homobiota were the only kind of life that could spontaneously arise, there is no reason that homobiotic life that is able to practice chemistry couldn't create heterobiota, tailoring the processes of their artificial life to the desired conditions. Eg: life that uses weird inorganic chemistry, life that doesn't use chemical processes, life that exists solely in silico.