Re: Lame philosphy

Friar Broccoli wrote:
John Wilkins wrote:
Noone Inparticular wrote:
John Wilkins wrote:

But in this case Salthe's trying for a teleological and
formalist notion of biology, which I do think is rather
lame. Hylomorphism (the idea that form is impressed upon
substance) and the corollary of final causation is, in my
view, a retrograde step. I think some people's
information-fetish in biology is a kind of hylomorphism.

I'm confused by what you meant with the comment in the
last sentence of this paragraph. Are you refering to
bioinformatics? Could you expound?

[... the following was unsnipped for better context ...]

But if we take the "informational" aspect of biology seriously
in the *right* way, we can see that what we are doing with
data is trying to find patterns that can be tested. There's
nothing wrong with this, so long as we don't hypostasise
(inappropriately treat as real) the data or the patterns.
Abstractions are in the head - in the world, we just have
physical things doing stuff, and they are often not what our
mental pictures suggest.

I'm going to take a run at the following to see how the
argument plays out. That is, I think I disagree:

Not exactly, but it can be. What I mean is the claim by people
like Dawkins that genes are information, as if that explained
something about them that isn't explained by simple (and
complex) biochemistry. But "information" is an abstract
property, not a physical one, and what genes do is because of
their physical properties.

I don't think "information" is abstract, although it is
not well defined. Second, I don't think that saying that what
genes do results from "their physical properties" is meaningful
without reference to information.

OK. Let's see how this goes...

As I understand things, a physical "property" is the result of
physical organisation; that is, it is a type of ordering, in the
negative entropy sense. At the deepest level, the primeval
stuff (whatever it is) is ordered into space, time, and matter
(say strings). Next up the strings are ordered into electrons,
quarks (neutrons and protons), gravitons photons etc. Up again
this ordering gives us atoms which differentiate into the
various elements. Then we see molecules as well as macroscopic
objects like planets, oceans etc (where entropy is usually
considered), and "above" that we get living organisms.

The very first sentence is, I think, wrong. Not because physical properties
aren't inherent in organised bodies, for they clearly are, but because the
notion of "order" here is entirely arbitrary. There is an exception I will
grant for the purposes of argument and because I'm an ignorant sod - a
thermodynamic property (exergy, is I think, the right term - negentropy is
Schrödinger's term and it is rather confusing; we are talking about the free
energy available in a phsyical system) is not arbitrary, but neither does it
correlate directly with "ordering", which is another abstract term.

In basic physics, in some ideal version of physics that gets it all right, we
may very well be able to find a unique descriptor of the fundamental elements
of the physical universe. But at higher levels of organisation, where the
elements of the universe are not constrained to one set of configurations, the
choice of metric used in "measuring" order is a matter of convenience. We may
choose to measure the electron bonds, or the size of the molecules, or the
cycling or the system in which those molecules exist, or different systems
that do not map uniquely on each other, and so on. As you get higher in the
(many) hierarchies of real systems, you get many choices about what to measure.

Biology realistically begins when exergy is employed to generate many
processes, and the states that this can attain are Many (capitalising to
indicate combinatorial complexity of a high order). So, where is the
"information"? It lies in the ways we describe and measure some elements of
the systems.

The point I want you to note here is that higher level objects
like atoms and molecules become "real" as a result of the
organisation of more primitive stuff. From here I want to argue
that this ordering is distinct from information content, or at
the very least that "information" is a *real* distinct class of
ordering that is not captured by *complexity*.

It is definitely related to the orderings. Information is that part of
complexity that we find salient and theoretically interesting.


Consider a gram of matter. As a mixed gas it is largely
unordered, however after baking in a sun and a bit of geological
kneading and sorting it might become a diamond of pure carbon, a
highly ordered (extremely low probability state). However, I
think most of us would agree that it doesn't contain a lot of
information. We could add information to it by systematically
modifying its lattice structure to encode messages. (We do
something similar when we disfigure a white *** of paper by
placing ugly black letters on it.) Note also that we could add
complexity by adding contaminating elements other than carbon,
which can improve the aesthetic properties of diamonds, but
this complexity is obviously not what we mean by information.

Note that the "information" that you are talking about here is not a measure
of complexity, but a semantic notion of information. If a lattice happened to
have an order that we could, if we had the appropriate lexicon, read out as
"VERY LIKE A WEASEL", that would not make the lattice an "information bearer"
unless that ordering was generated for the purpose of having the message.
Accidental orderings do not carry semantic information for that reason (one
reason why the "Bible Code" is meaningless).

In the formal sense, our diamond is even more highly ordered
after we have written a message in it, since its probability
state is now even lower. The point I am trying to make here is
that the ordering we see as a result of things like geological
processing does not give us the highest level of ordering.
Information is a distinct form of ordering that lies above
natural ordering and thus it is both real and distinct.

You define this, but I see no reason why we should take this fiat as being
"real" (herewith defined as that which exists independent of observers and
communicators).

In chemistry we see that very complex organic molecules can
form in liquid suspensions, or even in space. These may be
ordered/low-probability states, but they are clearly not in the
same class as those molecules when organised as genes, which are
read by the host organism as part of the process for making new
copies of the organism (including those same genes).

Point of order: question begging. Organisms do not "read" genes, they
transcribe (and reproduce) them in the appropriate conditions. If you call it
reading, fine, that's just a matter of technical jargon that has a resonance.
It's a very different thing to actually "reading". Using terms like "code",
"read", "copy", "edit" and the rest of the genomic jargon that is
informational is a convention designed to make it easier to teach the
concepts. It has no ontological import.

Thus living systems create order that is qualitatively different
(and not simply more complex) than the order arising from more
"primitive" natural processes. I believe this order can be
properly called "information" because it can be "read" and that
that information is most meaningfully (although not completely)
encapsulated in DNA. It may be possible to define information
in such a way that this type of organisation is not included,
but in that case you will need to coin a new word that lies
somewhere between "organisation" and "information".

More question begging. How is DNA "meaningful" (note that you shift again from
Shannon information to semantic information, two entirely distinct notions of
information)? If all you mean is that DNA is transcribed in such a manner that
the organism is able to survive to reproduction, then "information" here just
means a causal process. But if you mean something like "genes are information
because they "encode" phenotypic traits, then that is an artifact of the
language used, and is not evidence but a metaphor.

Thus, information, as higher level organisation is not an
abstract property, it is real (as real as molecules) with real
consequences.

The causal properties are real enough. But you haven't given any reason to
think that "information" plays any causal role here.


In bioinformatics one can treat information the right way, as
a pattern in data sets, or the wrong way, as a causal agent. I
have seen both done.

For the reasons described above I don't believe this sentence is
meaningful, because reading genetic data/information by the
organism is a significant part of the "cause" of that organisms
continuing existence. We may not be able to read that
information perfectly, but that doesn't change the information
into ordinary complexity, like that found in a weathersystem.

OK, let's consider a weather system. It is causally a process of
thermodynamics interacting with physical objects (molecules, mostly, except in
the case of lightning and static electricity). The weatherman represents this
in terms of measured data on a topographical map of the system under
investigation. The "information" is in the map, not the cloud system.

So how does this differ from organisms? The physical systems (which include
nucleotides) interact with each other in ways that generate some fairly stable
repeated states in the organisms, as well as a host of shorter-lived
molecules. They do all this in terms of thermodynamics; energy drives the
processes (and is shed as excess heat in every case). But "information"? It's
just the property of representations of symbols; A, G, C and T, or the
primary, secondary or higher conformational states of molecules. It's the map
that has the information, just like a weather system.

It's clearly of interest that some molecules are more exactly copied from
generation to generation, but the *causal* processes are all physical, and
information remains an abstraction in representations.

--
John S. Wilkins, Postdoctoral Research Fellow, Biohumanities Project
University of Queensland - Blog: evolvethought.blogspot.com
Servum tui ero, ipse vespera

.