On Wed, 9 Dec 1998 16:31:06, John McKechnie <J.M.McKechnie at ncl.ac.uk> wrote (in
Just addressing this part: The analogy you chose is probably not well-suited to your
> As an analogy we could think of two computer programs, one written for an
> IBM compatible PC, the other for an Apple Mac. It's conceivable that these
> programs could be exactly the same in terms of the binary string that
> encodes them,
Let's look at this backwards. Suppose you have two programs that "do" the exact same
thing, one on a PC and one on a Mac. The PC is defined as a machine with an Intel
microprocessor, typically running the DOS operating system (or some variant), and the
Mac as a machine with a Motorola or PowerPC microprocessor (typically running the
MacOS). The assembly-language instructions and hence the executable binary codes will
be different for the two programs, for two reasons. First, the microprocessors are
controlled by different assembly languages, whose binary representations are
different. Second, if the programs are written in a higher-level language, they will
make requests (calls) to the operating system, and these calls are different.
What about the opposite scenario, the one you propose? Is it possible for there to
exist a string of binary digits which corresponds to valid, executable instructions
for two different microprocessors? I think it possible but not plausible, because of
the conflicting constraints placed upon the code. As you point out, this might be
resolved by altering the hardware. Ultimately, this seems like a question for
theoretical mathematics: I bet one could prove that it is possible or impossible to
design such a bifunctional code.
> but because they are run on different computers, completely
> different behaviour is produced. The behaviour of the program arises from
> interaction of the binary string and the hardware. The behaviour of the
> program is not defined by the binary string in isolation.
True. One problem is that you have not defined what kind of behavior is acceptable.
Would you require that the instructions do anything useful? Probably not, since they
could evolve to have a function. You could tolerate an illegal instruction which was
ignored, but you probably wouldn't tolerate a series of bits which caused a system
> The program could
> be debugged by making changes to the hardware that interprets it as well as
> to the binary string.
In the example you started with, the "program" was to be readable by two different
microprocessors. Do you now allow it to run on just one processor? Or will you still
require that it run on both processors? That is, is there to be co-evolution of one
piece of software and two separate pieces of hardware? Possible, but certainly places
great constraints on the system.
> Similarly an organism's phenotype is determined by the
> interaction of its DNA and the environment in which it has developed.
The only sense that I can take this to be accurate is that the "environment" is the
set of macromolecules (enzymes, control proteins) and reactions (modification,
replication, decoding processes) which interact directly with a DNA molecule. This is
simply a consequence of the fact that DNA and the machinery for replicating and
decoding it have co-evolved.
On the broader front, the idea that DNA has no inherent meaning is both trivial and
nonsensical. Trivial, in that everyone agrees that if you (or a cell) didn't know
what to do with it, DNA would do nothing; nonsensical, in that we know in some detail
much of what DNA does (or has done to it) and how it is done. There is simply no
question about whether "another" nuclear environment could do something else with the
same piece of DNA: such "other" environments do not exist. The observable fact is
that the genetic code originated at about the time of the planet earth and has been
highly-conserved since then. Correspondingly, those proteins and RNAs examined which
decode DNA or RNA are also highly-conserved.
| Dr. Peter Gegenheimer | Vox: 785-864-3939 FAX: 785-864-5321 |
| Department of | PGegen at UKans.nospam.edu |
| Molecular Biosciences | http://rnaworld.bio.ukans.edu/ |
| & Dept. Evol Biology | |
| University of Kansas |"When you have excluded the impossible, |
| 2045 Haworth Hall | whatever remains, however improbable, |
| Lawrence KS 66045-2106 | must be the truth." S. Holmes |