River Out of Eden

River Out of Eden by Richard Dawkins (The Selfish Gene) gives an overview of our understanding of evolution. It explains deep concepts in understandable ways. Dawkins is a master in communication, and by using the ‘river out of Eden’ as an analogy, he presents evolution as a forwards flow of information. And although the book (and Dawkins in general) is a refutation of God-made creation(ism), it does the heavy lifting with explanation, not with conflict.


The ability to self-replicate is the (proximate?) cause for Darwinian selection, and the life we know on this world.

Chapter 1 – The Digital River

Real ancestors (vs myths/(religious) cults) hold the key to understanding life.

Ancestors are rare, descendants are common.”

Fun fact, not one of our ancestors died in infancy.

All organisms contain successful genes. Genes that have what it takes to become ancestors (to reproduce, leave kids behind). Genes to survive and reproduce.

Good genes cause success. Not the other way around (behaviour/lifetime doesn’t influence genes).

Every generation is a filter, only the successful genes get through. Some animals are sterile (worker ants), but they contain the genes that can also be passed along (the environment ‘chooses’ who becomes a reproducer or sterile worker). Thus they assist ‘their genes’ through the transgenerational sieve.

Genes are also not influenced by sex. Their effects are blended, but the genes are digital (yes/no, not analoge (radio frequency)).

The river analogy can be seen as genes travelling together on a stream. Those that cooperate well together, say in a body of an animal, form different branches/rivers. Speciation is the term for two rivers splitting. They will not join again.

The separation can be a geographical separation (both adapting to different environments over thousands/millions of years).

The number of species is estimated at 30 million (in 1994, now 2-10 million estimated to live), and if 99% has already gone extinct before now, the total branches/rivers (including those dried up) is 3 billion.

The separation of species (e.g. dinosaurs and mammals) may look significant, but it’s not. It was just another small river, branching from another. Only over long-history-time it looks significant.

The great animal groups are more similar in building blocks than we thought before. The genetic code is a dictionary with 64 words (from 4 letters) mapped on 21 words from another language (amino acids (20) plus punctuation mark). The chance of that is 1 in a million (x5). Or in other words, all life originates from a single ancestor.

So if you put on your molecular lens, all animals (and plants) are quite closely related.

DNA is digital, nerve cells are a mix of digital and analogue. The pulse (yes/no, action potential) is digital. But the rate of pulses is analogue.

This complex set of genes (and the instructions they give) are held together in a body (e.g. a polar bear). The number of cells of a polar bear are about 9 million million. And the complexity doesn’t stop there, each cell has a complex interior structure of folded membranes too.

Enzymes are the catalysts in a cell. Which genes in a cell are turned on, is determined by the chemicals already present in a cell. Bootstrapping is the term Dawkins uses for explaining how these processes start/interact. (do read the book or a whole book on this topic to get a better understanding of this).

[T]he genes that survive in the river will be the ones that are good at surviving in the average environment of the species, and perhaps the most important aspect of this average environment is the other genes of the species; the other genes with which a gene is likely to have to share a body; the other genes that swim through geological time in the same river.”

Chapter 2 – All Africa and Her Progenies

(cultural relativism bad)

Scientific beliefs are supported by evidence, and they get results [make testable predictions]. Myths and faiths are not and do not.

If we go back far enough in time, we are all cousins. If you go back to Roman times, the people there are either all our ancestors or ancestors of none (their line died out). Go further back and we’re all connected to the first replicator.

The changes in DNA can be measured with a molecular-clock (hypothesis, still somewhat controversial). The clock rate between species (and possibly time periods in history) may be different.

To find our common ancestor, we can look at mitochondrial DNA (because that doesn’t get mixed during sex, only that of the mother is passed along). Two million years ago is the moment of our mitochondrial (female line) ancestor (or as late as 250.000 years ago), probably in Africa.

Mithochondria are the powerhouses of our cells. If we look at their origins, they were bacteria (2 billion years ago).

… if all the mitochondria in a single human body were laid end to end, they would girdle the Earth not once but two thousand times.”

Chapter 3 – Do Good by Stealth

Creationists say something like “this is so beautifully designed, and it would be useless if it missed X (of X Y Z) function, God must have made this all in one go.” or as Dawkins puts it “… cannot have evolved by gradual stages, because the intermediate, half-formed stages could not have been good for anything.”

This chapter does away with those conceptions.

A proto-eye can already see (e.g. light and dark). Birds are fooled by red spots that their kin normally have. If you present a supernormal stimulus, they go crazy for it (as do humans, think adult movies).

Douglas Hofstadter (yes of Gödel Escher Bach) called the inflexible, mindless automatism that some (all?) animals exhibit (and bees in particular in this case) ‘sphexish’.

Many things we humans make also work when a part stops working (e.g. even a plane flies with one fewer engine). Something that breaks if it misses one part is called brittle (robust or antifragile could be the opposite?).

Eyes are useful in a gradient (analog) kind of way, you can vaguely see what is far away, and clearly see what is close. They have evolved between 40 and 60 times, with at least 9 different design principles.

Computer simulations show that an eye can be evolved (gradually) in about half a million years.

Do good by stealth. A key feature of evolution is its gradualness.” It may sometimes go quickly (e.g. meteor strike), but is almost always gradual.

The rest of the chapter describes the evolution of the dance of bees and some very clever experiments to test this.

Chapter 4 – God’s Utility Function

“Nature is not cruel, only pitilessly indifferent.

Nature is not cruel, only pitilessly indifferent. This is one of the hardest lessons for humans to learn. We cannot admit that things might be neither good nor evil, neither cruel nor kind, but simply callous – indifferent to all suffering, lacking all purpose.”

As humans we (think) we have a purpose, a goal, a consciousness. We plan for the future, look ahead, look back. But nature lacks this, nature just is. Evolution doesn’t have a plan. Evolution doesn’t answer the ‘what is it for’ question.

Only through Darwinian natural selection does evolution happen. There is no grand design or purpose. If we see that, it’s just an illusion left by the former.

Dawkins takes inspiration from Darwins Dangerous Idea (link if read). He uses the follow two terms:

  • Reverse engineering: making the assumption that there is an intelligent and economical reason for something being there (as outcome)
  • Utility function: that which is maximized

By watching the behavior of individuals throughout their lives, you should be able to reverse engineer their utility functions.”

There can be multiple things (utilities) that an organism (or organization for that matter) is optimising for. In the end, for us living things, it comes down to DNA survival.

Dawkins then explains the sex ratio and why a 50:50 division is optimal.

Beauty (e.g. peacock’s tail) is also explained by this utility. It isn’t directly useful for getting food, but displays evolutionary strength and over evolution it is selected for. Beauty has no virtue in itself, but the genetic competition makes sure it exists.

Evolution doesn’t have a ‘cooperative restraint’ in it. We can’t all just say, let’s not spend so much resources on beauty (or growing taller as a tree). Heck, the outcomes of this race could even mean the extinction of your species (e.g. all the beautiful birds get eaten by predators).

About old age and dying, Dawkins repeats some things I know about genes that optimise for reproduction, may be harmful if you’re older. They will not be filtered out (because with them, you can still get kids when you’re young).

About happiness. “Genes don’t care about suffering, because they don’t care about anything.”

If there is ever a time of plenty, this very fact will automatically lead to an increase in population until the natural state of starvation and misery is restored.”

Chapter 5 – The Replication Bomb

We are probably somewhere in space-time that can be called an information or ‘replication bomb’. Because life consists of replicators. And these replicators can lead to exponential growth.

This growth can only go on for so long, until more resources are acquired.

Dawkins then explains the start of this process, where self-replication differs from crystals (something building on itself, but not self-replicating).

The halfs need to split and then both sides need to be able to grow the other side again. DNA has four ‘letters’ that make this possible.

The copying isn’t perfect and because of how molecules can be folded, there is open-ended variety next to heredity.

Dawkins then describes several thresholds that a planetary replication bomb could/should pass:

  1. Replicator Threshold: self-copying system, with occasional random mistakes in copying. This leads to a mixed population with competition for scares resources.
  2. Phenotype Threshold: replicators survive because of causal effects on phenotype (parts of animals/plants that genes can influence).
    1. He has written more about this in The Extended Phenotype
  3. Replicator Team Threshold: working together in cells (eukaryotic cells is those in our body, otherwise bacterial cells which are the forerunners of them).
    1. Darwinian selection still chooses among rival genes, but the genes that are favored are those that prosper in the presence of other genes that are simultaneously being favored in one another’s presence.”
  4. Many-Cells Threshold: many cells working together to form a larger (emergent?) system (and that makes it different from crystals which is just molecules times X)
  5. High-Speed Information-Processing Threshold: neurons (at least on earth). This system needs sense organs, brains, and memory.
  6. Consciousness Threshold: humans, maybe other animals
  7. Language Threshold: networking system by which brains exchange information with sufficient intimacy to allow the development of a cooperative technology.
  8. Cooperative Technology Threshold: the meme, a river of culture
  9. Radio Threshold: sending out signals to outer space
  10. Space Travel Threshold: sending more than radio waves

Alright, that’s that.