Dear All,
In respect of the ongoing 'iota subscript' affair and the resulting
controversy surrounding the correct evaluation of John 1:1, Iain writes
as follows:
I would like to try to put this issue of the missing, or otherwise, iota
subscript in the text of John 1:1, into perspective.
All these objections ignore the fact that there is a coherent basis of
numerical evidence which ought to be assessed objectively as data,
independent of any preconceptions about what it "ought to be".
Experimental data is invariably noisy, but that does not stop us
inferring laws on the basis of measurement. We cannot tell if John
originally wrote in the iota or not; and that it seems to me is
irrelevant, and as irrelevant as pondering whether the scribe who copied
it in the Codex Sinaiaticus copied or did not copy an iota.
Let me illustrate this by enlarging on an excellent example of numerical
design given in Richard Dawkins's book "River out of Eden". I'll tell
the story in a slightly different way to Dawkins to illustrate the
point, but the principle is exactly the same. Dawkins sets out by
telling us that this is a perfectly feasible scenario in the not too
distant future.
Suppose at some time in the future, an infectious flu virus is sweeping
across the world, and we have advanced technology to a degree where we
can treat diseases by decoding the DNA base sequence, finding the active
proteins and develop drugs to order. The virus is therefore sequenced,
and someone notices, looking at a printout that in a non-coding part of
the DNA there is the following base sequence:
aaccttggaaatttcccgggaaaaaccccctttttgggggaaaaaaaccccccctttttttggggggg ...
and so on for a long time, a c t g cycling round with the number of
reps increasing each time. Counting the number of reps on each cycle,
you get 2,3,5,7,11,13,17,19,23,29, and you notice it is the first 10
prime numbers. After that, it is apparently junk DNA.
Dawkins asserts confidently that the sequence of primes would not occur
naturally, (i.e. did not arise through an evolutionary process), and
that we may therefore deduce that it was deliberately put there by an
intelligent mind. The purpose of it, he tells us, is to grab the
attention of someone, to draw attention - but to what? The biochemist
in the lab who spots this assumes it was put there by a human and that
it is a message in a bottle ( "sneezed around the world" as Dawkins
brilliantly puts it). A reasonably simple decoding scheme is applied to
the codes that immediately follow the prime sequence, and they find a
message in English from a biochemist who has been kidnapped by a
despotic regime, and is being forced to use his skill in developing
deadly viruses for germ warefare.
But now let us vary the story a little. Suppose by the time it reaches
the lab, the virus has mutated a little, and supposing the seventh base
on the string is now a 't' instead of a 'g'. Now it doesn't make a
perfect sequence of primes. But how many people would then say that
because of a single mistake that has crept in, that the whole sequence
is thereby invalidated? I suggest no-one; the pattern would still be
evident if, say 1% of the codes were mutated, maybe even 5%. Perhaps by
the time 10% got corrupted one might begin to doubt that it was anything
more than coincidence. But in the case of the missing iota, we are
talking about a single change. With the iota, multiple instances of
interlocked numerical phenomena are observed, just as indicative of
numerical design as the example in Dawkins's book.
To dismiss the whole thing on the basis of one possible "mutation" is,
it seems to me no different from dismissing the evidence on the basis
of a single point mutation in the Dawkins example. No respectable
scientist would ignore the observed underlying pattern on the basis of
such a thing.
To briefly address another point; let me re-iterate - the numbers are
not important in themselves; but they are an attention-grabbing device
to suggest that the text was produced (inspired by, designed, call it
what you will), by an external intelligent source.
Regards,
Vernon
This archive was generated by hypermail 2b29 : Thu Jul 12 2001 - 17:55:23 EDT