Definition of Life

[Julian Blanc]

Ewan birney wrote:

: Julian Blanc wrote:
: > The best definition I heard was given by John Maynard Smith. A living 
: > thing must a physical dissipative structure having the following 
: > characteristics:
: > 
: > 	- Variation 
: > 	- Reproduction
: > 	- Heredity
: > 

: 	Again I would like to use the example of the chain letter: it 
: satisfies the above criteria. (by chain letter I mean the things telling
: you to write it on to the ten more  people else the recipient falls foul
: of terrible luck etc). It has a physical structure, it reproduces (albeit
: using humans as an unwitting host...but so do many, many parasites which
: are very much alive), it has variation and hereditary. So John Maynard Smith
: would call chain letters alive...

: 	So...is a chain letter alive, and if it is not, what criteria is
: it not satisfying?

Although chain letters are physical structures, they are not dissipative. 
There is no flow of energy or particles through the physical structures 
which act as vehicles (the letters themselves).

Now, this might indeed be a practical definition. I see nothing wrong 
with that, when whole branches of biology, such as taxonomy, are
based on practical divisions and definitions. Practical definitions have 
the property of conferring false peace of mind, but peace of mind after all. 

We could say that chain letters, and why not, computer viruses and 
cellular automata, are bloody good simulations of life. In some cases the 
similarity might be so great as to render it practically indistinguishable 
from a "truly living" creature. But just because a Viroy buterfly is 
practically indistinguishable from a Monarch, it doesn't mean that they 
are the same species. Practical definitions are handy, especially when we 
have a gut feeling that they are right but we can't say why. Maynard 
Smith's definition is the closest and most elegant explanation to that 
gut feeling.


Julian Blanc