UP

Notes by RBJ on

Darwin's Dangerous Idea

by Daniel C. Dennett

Part I - Starting in the middle

Chapter 1Tell me Why
Chapter 2An Idea is Born
Chapter 3Universal Acid
Chapter 4The Tree of Life
Chapter 5The Possible and the Actual
Chapter 6Threads of Actuality in Design Space

Chapter 1 - Tell me Why

This is mainly historical background, motivation, context setting. Locke and Hume are singled out as pre-Darwinian philosophers who held that mind must have come before, and designed, the material universe (whether Dennett is correct in attributing this view to Hume is not particularly important).

Chapter 2 - An Idea is Born

First the basic ideas of Darwin's theory are presented. Evolution, natural selection, and a discussion of whether Darwin did or didn't explain speciation. Then the idea of evolution as an algorithmic process is introduced, followed up by a more general discussion of processes as algorithms. At this stage I'm in the dark, not only about why Dennett thinks its a good idea to think of evolution as an algorithm, but even about exactly what he means by that. Its not clear to me why the idea of algorithm is preferred to the broader notion of mathematical model, or why he says that evolution is an algorithm rather than can be modelled by or can be simulated by an algorithm. Its also not clear how he is to reconcile the non-determinism which he seems to find important with the notion of turing computability which he appears to be working with.

Chapter 3 - Universal Acid

The name of the chapter tells it all. This is where Dennett starts to get seriously into gear telling us that Darwin's idea is really a bit of a skyhook. (That isn't what he says, "skyhook" is a term of abuse)

Now I always thought that Darwin's idea was just about evolution of living things, and that the big thing in his story was natural selection. Dennett is painting on a much larger canvass here, and for someone like myself, who hasn't actually read Darwin, its hard to tell how much of this is Darwin's idea and how much is generalisation by Dennett.

So we start off with this "cosmic pyramid" something like this:
God
Mind
Design
O r d e r
C h a o s
N o t h i n g
according to philosophers before Darwin (according to Dennett) the universe starts at the top and progresses down the pyramid. Darwin turns it upside down. We start with nothing, progress to chaos, then things get to be a bit more organised and eventually life evolves with the kind of intricate structure we associate with design. Life evolves into intelligent life, so mind has appeared, and the next step beyond that presumably involves something really magic (maybe Darwin's dangerous idea).
Not only do we have a progression here, we have an accelerating progression. The evolution of sexuality makes a radical difference to the pace at which evolution can take place. More adaptable organisms also evolve faster (the Baldwin effect), so intelligence, as it evolves also speeds up the process of evolution.

It also looks like the kind of things which one might have thought spelled a later limit to the scope of Darwin's theory, e.g. when medical technology enables the unfit to live and breed, and genetic engineering enables us to design our own children, that this is just another thing which fits into the same frame and just makes things wheel along even faster.

So I've lost track of how something can fail to be evolution, has Dennett just generalised the concept out of existence?

Chapter 4 - The Tree of Life

The tree of life is probably one of Dennett's intuition pumps. Dennett is really working hard here to present his material in a graphic and intelligible manner, but I guess I'm not his target audience - it didn't really work for me.

We get the overall story, with the major milestones:

prokaryotes - bacteria, single cells without nuclei
eukaryotes - single cells with nuclei
multicellular organisms
sexual repoduction
each of which results in an acceleration of pace of evolution.
We also have this idea that as we progress through these stages we are getting "more design", but I never managed to figure out what he means by that. (design is not something that I think of as quantitative)
Big deal about speciation, but I can't figure out Dennett's position. On the one hand he seems to think that essentialism is as dead as a dodo. On the other, he is unable to offer any solid definition of the word species (nothing solid either in the reality or in the language) That seems to me to leave speciation as just a matter of taxonomy. But he talks as if he thinks that species boundaries are real rather than discretionary. He makes mystery of the fact that the places (on "the tree of life") where the species transitions takes place are not known until long after they have happened. I can't see what the mystery is, we don't know where they happen until later, because we didn't decide where to draw the line until later!

If you aren't an essentialist and can't come up with a solid definition of species, then to understand speciation you have to look, not at the tree of life, but at the institutions which regulate how species are named. Gould's essay "Bully for Brontosaurus" is more apposite than Dennett on this topic.

Chapter 5 - The Possible and the Actual

I think Dennett, though a philosopher, has a bad case of Biologist's PE (Physics Envy). He's looking for biological laws, and universal ones at that. We can't have it that biology is just taxonomy.

The search for fundamental bioligical laws is connected with biological possibility in the same way that the notion of physical possibility is closely coupled with the laws of physics ("physically possible" means "consistent with the laws of physics"). So talk about possibility here is neat, because its either philosophical or scientific, depending on whether you slip in the qualifier biological.

Dennett's big idea in this chapter is that biological possibility can be defined through the idea of accessibility in the genome space. This seems to me an interesting idea, and from a philosophical standpoint its the bit of the book I like best, so far.

Even so, I do find it all a bit loose.

Chapter 6 - Threads of Actuality in Design Space

There is one design space. This seems to be the culmination of Part I.

Well, I've come to the end of part one hearing the message, but not having much clue what it means. I can think of several possible meanings under which it is a truism, but have not acquired any sense in which it is both true and significant.

If, as Dennett seems to believe, design is all about algorithms and their results, then the claim is a truism, though the premise is a devastating empoverishment of the mathematics on which science has hitherto been based.

If we take a broader view of design, still distinguishing a design from the thing of which it is a design, a design might be construed as a mathematical model, or a formal specification. In this sense (which is not Dennett's) there is again arguably just the one design space, and the point is not so obviously a truism. But it's logic, not biology.

I'm afraid I just don't get the point. Dennett seems to be talking about the genome space, but this does not compute. It does not have the design of my PC in it, or even my hammer. Is he just talking about biological design? I think not, the tone is cosmic.


UP HOME © RBJ created 1997/1/12 modified 1997/2/3