Time to update taxonomy? - A thought experiment.

Roger Hyam roger at HYAM.NET
Sat Feb 1 23:33:18 CST 2003


The DNA taxonomy arguments raised recently have been exercised to death
over the past few years both on this list and in the published literature.

So here is my, probably unoriginal, and rather dull contribution to the
argument. I apologise in advance if you have heard it all before.

Lets do a thought experiment. Imagine you were a very talented molecular
biologist and you worked so quickly that at any one moment in time you
had the entire DNA sequence of all the individuals on the planet on the
hard disk of you computer. That is all the *individuals* every instance
of every organism and every last bit of DNA (or RNA if you like)
sequenced down to the last base pair. (We will skip the difficulty of
defining what an individual is as this is a thought experiment. On
similar grounds we will also skip the technicallity of how we manage to
sequence everything without grinding it all up - just image we can do
it). Your laptop is a Dell SuperDuper running Windows ZP and is
therefore infinitely powerful and can look at all possible trees in a
trice ( np-complete == no problem ). You can hit enter and it will
reveal to you the one true tree of life and all the species boundaries.

Do you think the last sentence of our thought experiment is reasonable?
  What assumptions do we make here?

Firstly we assume that there will be only one most parsimonious tree (
or one undisputable topology produced by whatever method )when infact
the tree has to be reticulate at the top because so many organisms are
sexual and so do not have simple branching ancestries let alone the
problems associated with homoplasy, clinal variation, hybridisation,
active speciation, organelle inheritance etc etc. The totality of
biodiversity can not be described in terms of a simple branching tree.
If you believe it can then your belief is probably a religious one. So
when we hit enter we are likely to get a curates egg of a tree. Good in
parts but more or less confused towards the top (i.e. the important bit
- I think ) We also assume that we can define what a species is which we
can't - ask an under graduate student to write and essay on it if you
have any doubts. No amount of DNA is going to help us sort that one out
although we may be able to make up some arbitrary rules that only offend
half the taxonomists.

Our thought experiment has shown that in looking to DNA to solve the
major problems of taxonomy we are seeking technological solutions to
intellectual problems. No more saying "If we just sequence another gene"
or "If we could just sequence a few more individuals". Even if we had
all the data and all the computing power we still wouldn't answer the
question.

I presume the goal of taxonomy/systematics is to produce a set of units
arranged into some form of system that we can all agree on that reflects
something or is useful for something. How/Why/What we do is an
intellectual question not a technological one. You can throw as much
data at it as you like you won't find and answer. You have to think
about it.

If taxonomy is backward it is in its conceptual approach not in the
technology it uses.

The answer you would get if you sequenced everything and gave it to an
infinitely powerful computer is, of course, 42. Trouble is we are too
lazy to work out what the question really is. Perhaps, if we did, people
would be more likely to fund us to answer it.

You can all flame me now!

Roger




More information about the Taxacom mailing list