On Mon, 10 Feb 1997 21:42:27 -0500, Brian D. Harper wrote:
>BH>I don't mean this to be negative, my suggestion is that you just
>let slide stuff more than a week or two old and get caught up.
>You'll be much more effective this way.
>SJ>OK. I might do that.
I haven't unsubscribed, but I am catching up by only responding to
posts (with one or two exceptions) that have my name in them. After a
bit of a lag, this is now working. I hope to post every couple of
days.
BH>To help you out, I've decided not to reply to any of your most
recent group of posts except for this one.
That's fine by me. Neither Brian, nor anyone else need feel that they
have to reply to my posts or even read them! They are addressed to
the Group and I know from private messages I get that some lurkers
find them useful.
BH>I chose to reply to this one for four reasons (1) information and
>complexity is one of my favorite topics (2) unlike most of our other
>"discussions" this one may have some interest to everyone else (3)
>this topic is, IMHO, extremely important to the debate on origins and
>(4) Bill Dembski's NTSE conference paper "Intelligent Design as a
>Theory of Information" could prove very useful in getting over the
>hurdle of defining what specified complexity is.
>
>Burgy provided us with the URL for this:
>http://www.dla.utexas.edu/depts/philosophy/faculty/koons/ntse/ntse.html
Thanks to Burgy for this. I have looked up this site and it looks
interesting. Some of the papers are already posted under:
http://www.dla.utexas.edu/depts/philosophy/faculty/koons/ntse/papers/
>BH>There were a couple of reasons for my challenge above. One was
to see if you had any understanding of the quotes you were giving
>out.
There is often a patronising tone to Brian's posts to me, but this is
particularly so.
[..]
SJ>...I will stick to laymen's definitions like "specified
>complexity".
>SJ>"Information in this context means the precise determination, or
>specification, of a sequence of letters. We said above that a
>message represents `specified complexity.' We are now able to
>understand what specified means. The more highly specified a thing
>is, the fewer choices there are about fulfilling each instruction.
>In a random situation, options are unlimited and each option is
>equally probable." (Bradley W.L. & Thaxton C.B., in Moreland J.P.
>ed., "The Creation Hypothesis", 1994, p207)
[...]
>B&T's statement: "In a random situation, options are unlimited and
>each option is equally probable."
>
>I followed up on this oops later in this thread in an attempt to
>clarify, you probably haven't seen it yet.
There is nothing wrong with this definition. It is *exactly* what
"random" means:
"Perhaps the most important sample type is the random sample. A
random sample is one that allows for equal probability that each
elementary unit will be chosen...Random numbers are digits generated
by a process which allows for equal probability that each possible
number will be the next." (Lapin L., "Statistics for Modern Business
Decisions", 1973, pp194-195)
Brian appears to be getting mixed up with the *Darwinist-biological*
definition of random as in "random mutation" (which indeed does not
mean that "each option is equally probable"), but B&T are not talking
about "random" as in mutation.
BTW, Brian, just skipped over the main point which was:
"Information in this context means the precise determination, or
specification, of a sequence of letters. We said above that a
message represents `specified complexity.' We are now able to
understand what specified means. The more highly specified a thing
is, the fewer choices there are about fulfilling each instruction.
(Bradley W.L. & Thaxton C.B., in Moreland J.P. ed., "The Creation
Hypothesis", 1994, p207)
Perhaps he would care to comment on Bradley & Thaxton's
definition of "specified" = "fewer choices"?
[...]
>SJ>This is *not* an example of increasing the information of an
>already specified complex string. The string of AAAAs has zero
>information content, so anything would be an improvement! But this
>has *no* analogy with a living system. There may indeed have been
>an increase in "algorithmic information content" by a "random
>mutation" but I cannot see that it has "created" any "new
>information", in the sense that I am using it, ie. on the analogy
>of an English sentence like "John Loves Mary"....
>
>"...Yockey and Wickens develop the same distinction, explaining that
>"order" is a statistical concept referring to regularity such as
>might characterize a series of digits in a number, or the ions of an
>inorganic crystal. On the other hand, "organization" refers to
>physical systems and the specific set of spatio-temporal and
>functional relationships among their parts...." (Thaxton C. ...
>,"The Mystery of Life's Origin" 1992, p130)
>
>Maybe Brian can do the above with a *real* English sentence, like
>THIS SEQUENCE OF LETTERS CONTAINS A MESSAGE?
BH>I will do just that in a moment. It is important to emphasize at
>this point though that dealing with English sentences is just by
>way of analogy and all analogies break down eventually. For
>example, in English there is a lot of intersymbol influence whereas
>in proteins there is no intersymbol influence (I'm not an expert in
>molecular biology either :) so someone please correct me if I'm
>wrong).
This is true of English, but this would only affect the statistical
structure, not make it impossible:
"The statistical structure will be reflected by the fact that some
symbols may be chosen more often than others. Some symbols may
often, or always, be followed by certain others. That is, there may
be an intersymbol influence which may extend some distance along the
series. For example, in English q is always followed by u, "ght",
"th", "ch", "ph", etc., often occur together." (Yockey H.P., "An
Application of Information Theory to the Central Dogma and the
Sequence Hypothesis", Journal of Theoretical Biology, 46, 1974,
pp371-372)
In any event, the real analogy is with human languages, not just
English, and they might have less "intersymbol influence" than
English.
BH>But the real problem with the English analogy is the temptation to
>draw meaning from the words themselves. This is a confusion that
>plagues so many discussions of information on both sides of the
>fence. Dawkins falls for this trap in his "me thinks it is like a
>weasal" word game. More subtly, Manfred Eigen also commits the
>error with his "value parameter" in his Hypercycles scenario for
>the origin of information.
Indeed. The word "information" as defined by information theory does
not deal with meaning:
"Some have attempted to develop a theory of meaning from these ideas.
Shannon (1949) warned against this at the outset of his paper. The
assignment of meaning or in biology, specificity, to certain
particular member of the ensemble lies outside information theory."
(Yockey H.P., "An Application of Information Theory to the Central
Dogma and the Sequence Hypothesis", Journal of Theoretical Biology,
46, 1974, pp371-372)
which is the whole point of "information" in biology and human
languages. That's why I don't accept Brian's "information theory"
definition of "information" and prefer to use "specified complexity"
instead.
BH>The pioneers of information theory warned of this trap from the
>beginning:
> ================================
The fundamental problem of communication is that of reproducing at
one point either exactly or approximately a message selected at
another point. Frequently the messages have MEANING; that is they
refer to or are correlated according to some system with certain
physical or conceptual entitites. These semantic aspects of
communication are irrelevant to the engineering problem. --Shannon,
<Bell System Technical Journal> v27 p379 (1948).
================================
Yes. See above. Information theory is concerned with "the engineering
problem" of "reproducing at one point either exactly or approximately
a message selected at another point". It has nothing to do with the
creation of the meaning in the first place.
BH>To tie this in with biology we can observe that the genetic
>information processing system can process the information for
>forming a non-functional protein as easily as it can for a
>functional protein.
I am not so sure that this is compeletely true. Some "non-functional
proteins" (eg. D-amino acids) may be unable to be processed by "the
genetic information processing system."
BH>Another information pioneer said something very similar:
>
> ===================================================
>What I hope to accomplish in this direction is to set up a
>quantitative measure whereby the capacities of various systems to
>transmit information may be compared. Hartley Bell System
>Technical Journal v7 p535-563 (1928).
===================================================
Yes. Again, this is to do with "the capacities of various systems to
transmit information", not the problem of the meaning of the
information.
BH>I note that your quote of TB&O above refers to Hubert Yockey and
>Jefferey Wicken. There is good reason to mention these two since
>both are recognized experts in biological applications of
>information theory. Its been awhile since I've read Wicken so I
>won't try to rely on my memory to discuss his views. Yockey I'm
>much more familiar with. Everything I've written here follows
>directly from Yockey's work. Let me give a few quotes.
>
>First, the one Glenn likes to quote so often ;-)
>
>====================================
>Thus both random sequences and highly organized sequences are
>_complex_ because a long algorithm is needed to describe each one.
>Information theory shows that it is _fundamentally_ _undecidable_
>whether a given sequence has been generated by a stochastic process
>or by a highly organized process. -- H.P. Yockey, _Information
>Theory and Molecular Biology_, Cambridge University Press, 1992, p.
>82.
>=====================================
>
>Note that *both* random sequences and highly organized sequences
>are complex (contain a lot of information).
Again, the word "information" is being used in a different sense from
what it means in biology and human languages.
>====================================================
>The entropy that is applicable to the case of the evolution of the
>genetic message is, as I believe the reader should now be convinced,
>the Shannon entropy of information theory or the Kolmogorov-Chaitin
>algorithmic entropy. ...
>
>The Kolmogorov-Chaitin genetic algorithmic entropy is increased
>in evolution due to the duplications that occurred in DNA. [...]
>Thus the genetic algorithmic entropy increases with time just as
>the Maxwell-Boltzmann-Gibbs entropy does. Therefore creationists,
>who are fond of citing evolution as being in violation of the
>second law of thermodynamics (Wilder-Smith, 1981; Gish, 1989), are
>hoist by their own petard: evolution is not based on increasing
>_order_, it is based on increasing _complexity_.
I am not aware that "Wilder-Smith" or "Gish" actually say that
"evolution" is "in violation of the second law of thermodynamics".
I would invite Brian to post where he or Yockey claims they do. In
his chapter "Creationist Theory: Popular Evolutionist
Misunderstandings", Ratzsch says:
"Perhaps the most prevalent of the misconstruals of creationism
involves the Second Law of Thermodynamics." (Ratzsch D.L., "The
Battle of Beginnings", 1996, p91)
BH>In fact, evolution requires an increase in the
>Kolmogorov-Chaitin algorithmic entropy of the genome in order to
>generate the complexity necessary for the higher organisms. Let us
>recall from section 2.4.3 that _highly organized_ sequences, by the
>same token, have a large Shannon entropy and are embedded in the
>portion of the Shannon entropy scale also occupied by _random
>sequences_.Evolution is not in violation of the second law of
>thermodynamics. This is what any reasonable scientist believes;
>nevertheless, it is important to avoid word-traps and to reach the
>correct conclusion for the correct reasons. -- Hubert
>Yockey,_Information Theory and Molecular Biology_, Cambridge
>University Press, 1992, p. 310-313.
>===========================================
[...]
>SJ>The above is not an "example" at all. And I find it strange
that I am referred to a web site. I do not regard web sites as
"the scientific literature".
BH>Well, first of all, the web site you were referred to contains
>almost all of the papers (published in the "best" journals) of one
>of the founders of algorithmic information theory. A real time
>saver, especially since many are not going to have ready access to
>most of the journals. In any event, what you were referred to is
>definitely the literature.
OK. But I am not sure from my discusions with Brian to date that
"algorithmic information theory" is the same "information" that
I was talking about when I said: "Sorry, but one thing that `random
mutation' cannot do is to `create new information'"
BH>More importantly, this was not the example I was talking about.
>Note that I referred you to the "pure chance" thread where this
>example was discussed extensively. Since you have been skipping
>over messages that don't have your name in it you probably missed
>the example:
>
>J.S. Rao and C.P. Geevan, "Significance of the Information
>Content of DNA in mutations and Evolution," <J Theor. Biology>,
>96:571-577, 1982.
Again, thanks to Brian. But again, I think that, from the examples
that Brian has used from information theory, that the "Information"
used here is not the same "information" that I am referring to.
[continued]
God bless.
Steve
-------------------------------------------------------------------
| Stephen E (Steve) Jones ,--_|\ sejones@ibm.net |
| 3 Hawker Avenue / Oz \ Steve.Jones@health.wa.gov.au |
| Warwick 6024 ->*_,--\_/ Phone +61 9 448 7439 (These are |
| Perth, West Australia v my opinions, not my employer's) |
-------------------------------------------------------------------