|
Author
|
Topic: Objective vs. Subjective Information
|
James A. Barham
Member
Member # 50
|
posted 19. March 2002 08:09
At the risk of becoming a bore on the subject, I would like to take another stab at generating some discussion on what I take to be the crucial point at issue between the ID and the self-organization perspectives.
I view the concept of information lying at the heart of the ID approach, and especially Dembski's notion of the conservation of information, as problematic due to its inherently observer-dependent and hence subjective character.
Now, all human observations are obviously observer-dependent in one sense: it is we who make them. And all information is objective in another sense: everything that we do also exists, has its own objective reality. However, when one organism (a human) observes another organism (say, a bacterium), we can make a further distinction. Say that the first organism H observes a certain pattern P in the second organism B. The question is, Does P have an intrinsic functional importance for B, or is P just an artifact of H's arbitrary choice of what to pay attention to?
The Shannon-Kolmogorov-Chaitin-et al. information-theoretic or algorithmic approaches (the distinctions here are unimportant for my point) have no way of making this distinction. P will have a certain information content regardless of its functional value for B. It is true that this information content will be objective in the sense that it is formalizable and reproducible, and hence subject to consensus among all H observers. However, the information-theoretic interpretation of P is inherently subjective in that in itself it says nothing whatever about the functional value of P for B itself.
Now, I am a realist. I believe that teleology, function, and value are real biological phenomena in no way dependent on the human observer. I also believe that the concept of information has a crucial role to play in biology. But I submit that in order to understand the functional or semantic value of P for B itself, we require an entirely different approach to understanding the phenomenon of biological information. It will still be a theory devised by us, of course, on the basis of H's observations of B, but it will be a theory that describes information as an objectively existing biological phenomenon---defined by its value for B---and thus something that enjoys exactly the same ontological status as all the other categories we find in (and do not merely impose on) the world.
(For anyone interested, I have presented a speculative alternative model of biological information as a dynamical process in "A Dynamical Model of the Meaning of Information," BioSystems, 38: 235--241.)
IP: Logged
|
|
John Bracht
Member
Member # 5
|
posted 19. March 2002 12:25
James raises some good points. I have spent a fair bit of time thinking about how to apply Dembski's concept of specified complexity to biological systems, in such a way that we are not just seeing what we want to see. I think Dembski has given us some clues about how that might be done:
quote:
Biological specification always denotes function.
( Intelligent Design, Illinois:IVP. 1999. p 149.)
James comments,
quote:
The question is, Does P have an intrinsic functional importance for B, or is P just an artifact of H's arbitrary choice of what to pay attention to?
I think this is an important question, and I submit that the answer is: P has intrinsic functional value for B, and we can show it in the laboratory.
Biological systems demonstrate remarkable functionality. Many proteins function as enzymes, catalysts for reactions (catalysts help a reaction go faster but are themselves unaltered by the reaction). Their rates of catalytic enhancement are measurable in the laboratory, and they outperform any other known catalysts by orders of magnitude. B. W. Matthews, C. S. Craik, and H. Neurath, in an article titled "Can Small Cyclic Peptides have the Activity and Specificity of proteolytic Enzymes?" (PNAS, vol 91, 4103-5), describe trypsin (a serine protease) the following way:
quote:
The primary binding site of trypsin is a cavity that is approximately 10 Å deep and 4 Å x 6 Å in cross section with an aspartic acid located at its base. When a peptide substrate binds to the enzyme the binding pocket not only provides specificity by preferentially binding an arginine or lysine side chain but also helps align the scissile peptide bond relative to the catalytic triad [a characteristic set of 3 amino acids that actually do the work of catalysis] and the oxyanion hole [a feature that helps align the substrate and stabilize the transition state during catalysis]. It is this synergistic effect that contributes to rate enhancement of nine orders of magnitude for the serine proteases when the mechanisms of enzyme and non-enzyme-catalyzed reactions are compared. [emphasis added]
My biochemistry text lists the following characteristics of biological catalysts (enzymes):
quote:
1. Higher reaction rates. The rates of enzymatically catalyzed reactions are typically 10^6 to 10^12 times greater than those of hte corresponding uncatalytzed reactions and are at least several orders of magnitude greater than those of th ecorresponding chemically catalyzed reactions.
2. Milder reaction conditions. Enzymatically catalytzed reactions occur under relatively mild conditions: temperatures below 100C, atmospheric pressure, and nearly neutral pH. In contrast, efficient chemical catalyzsis often requires elevated temperatures and pressures as well as extremes of pH.
3. Greater reaction specificity. Enzymes have a vastly greater degree of specificity with respect to the identities of both their substrates (reaqctants) and their products than do chenmical catalysts; that is, enzymatic reactions rarely have side products.
4. Capacity for regulation. The catalytic activities of many enzymes vary in response to the concentrations of substances other than their substrates. The mechanisms of these regulatory processes include allosteric control, covalent modification of enzymes, and variation of the amounts of enzymes synthesized.
(source: Voet and Voet, Fundamentals of Biochemistry. New York: Wiley and Sons. 1999. p. 282-3)
In the same text, on page 539, describing the photosynthetic reaction center of purple photosynthetic bacteria, the authors note
quote:
In fact, electron transfer in the reaction center is so efficient that its overall quantum yield (ratio of molecules reacted to photons absorbed) is virtually 100%. No man-made device has yet approached this level of efficiency. [emphasis in original]
Biological enzymes have rate enhancements of 6-12 orders of magnitude and reaction efficiencies that are amazing, often 90 or nearly 100% efficiency. Furthermore, many poisons affect enzymes and thereby "gum up" their function, with the resulting death of the organism. In other words, this high functionality of biomolecules seems to be more than a luxury; it appears to be a requirement for life to combat the constant tendency for thermodynamic breakdown. Thus, life is highly constrainted by independently given functional "targets." The high degree of functionality exhibited by these systems suggests that the "target" is very small relative to the rest of sequence space. I suggest that these independently given functional patterns are the specifications described by Dembski as legitimate patterns for detecting design.
What is the relationship between this extreme biofunctionality and the concept of information? Information always denotes some sort of signal agreement between two parties. The meaning of a symbol is arbitrarily assigned and agreed upon by the two parties; when the message is sent by party A it can be read by party B. The same thing happens in the cell. The "meaning" of the message (gene) is determined by the function of its protein. The fact that a gene "means" any particular protein, just as in ordinary language, is fully arbitrary, and is mediated by an arbitrary assignment of values in the genetic code. The word "car" could correspond to anything we want it to, but we have chosen to apply it to motorized vehicular devices (with certain characteristics). In the same way the correlation between a genetic sequence and its corresponding protein is arbitrary and mediated by the physical instantiation of the tRNA's and ribosome machinery following the genetic code. Rather than being purely an abstract connection, genetic meaning is mediated through physical operators from message to function. However, the principle is, I think, precisely the same.
I don't know if this helps, but it's my stab at an answer to James's question. Any comments?
John Bracht
[ 19 March 2002, 12:31: Message edited by: John Bracht ]
IP: Logged
|
|
Janitor@MIT
Member
Member # 125
|
posted 19. March 2002 18:15
I would characterize the problem appositely: The statistical measures effectively eliminate the observer-dependency by rendering information trivially "objective." By these measures the information content of a DNA molecule on the back of a licked stamp is the same as that contained in a fertilized ovum. In the words of Schrodinger (?) it doesn't tell us the difference between a live lion and a dead dog.
It is exactly in the observer-dependecy of information that life can be distinguished from physics and chemistry. I think this is an important distinction that cannot be minimized or eliminated.
I think John Bracht has touched upon the problem/solution: the "semiotic" content is related to function (which I would further relate to control). Circularly information is meaningfully functional and functionally meaningful.
Reinvesting life with meaning is a worthy cause and I am encouraged that individuals worthy to the task have taken it up. I am particularly distressed that evolutionary biology is not merely artless science (a frequent complaint of mine), but this dysteleological outlook serves as the philosophical "warrant," by way of a sort of self-fulfilling prophecy, for the nihilism that humans are in any case susceptible to.
Is meaning as much a part of the universe as gravity, space, and light? Is it a question of reducing "meaning" to function? I don't have a problem with that (although as stated it looks worse than it is, I think).
I like James Barham's equation to prediction and I think it is closely related to John Bracht's implicit argument to control. A synthesis may be in order here.
Anyway, rather than rambling on endlessly (my usual policy) I'll leave it at that.
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 19. March 2002 22:57
Thanks for your comments, John and Janitor.
I don't disagree with anything either of you said, but I think the main point of contention is still not quite fully in view.
The problem is, What is the ontological status of the functional organization of the cell? If it is objectively real, then how can we account for it?
One option is Howard Pattee's "complementarity" position: There is a mechanistic account and there is a teleological/semantic account, and neither is reducible to the other. Like "compatibilism" between free will and determinism, I guess I find such "solutions" temperamentally unsatisfying. It seems to me that there is a genuine scientific problem here: Namely, how to account for the massive coherence and coordination of all cellular processes with the ultimate goal of self-preservation of the whole. Natural selection does not help here, because it simply presupposes the functional organization of the cell (which is implied in the very word "survival," not to speak of "fitness," "success," etc.).
I don't pretend to have an answer, but it does seem to me that if an answer lies anywhere in our current knowlege (albeit in nuce), then it must be in the sciences that study coherent and collective phenomena: to wit, nonlinear dynamics and condensed-matter physics.
IP: Logged
|
|
John Bracht
Member
Member # 5
|
posted 20. March 2002 00:14
James,
I am genuinely trying to get a clear view of precisely what you are getting at in your original and last post. I would like to come to understand precisely what it is you are arguing. However, I'm having trouble pinpointing it from the last post so I'll quote it and ask some questions.
quote:
The problem is, What is the ontological status of the functional organization of the cell? If it is objectively real, then how can we account for it?
As a biology student, I can say with certainty that the functional organization of the cell is ontologically real, at least as real as anything else we determine by scientific investigation. Is this really in doubt? One of my biology professors once told me "We all see the Hand in biology." The functional organization is literally everywhere in biology. It permeates it. In fact, one could say that biology IS functional organization. So I honestly don't see how the ontological status of this functional organization could be in doubt. But perhaps I misunderstand the question? Please clarify for me.
quote:
t seems to me that there is a genuine scientific problem here: Namely, how to account for the massive coherence and coordination of all cellular processes with the ultimate goal of self-preservation of the whole. Natural selection does not help here, because it simply presupposes the functional organization of the cell (which is implied in the very word "survival," not to speak of "fitness," "success," etc.).
Ok, now it seems we've changed topics. Rather than arguing about whether the functional organization is real, you are arguing that natural selection cannot account for that organization (implicitly assuming that it is, indeed, real). Here I definitely agree with you. But again, I want to make sure I am understanding your point. Am I?
Just trying to get things clear.
John Bracht
P.S. Janitor, I like your comment about functionality and how it relates to control. You're absolutely right about this; most cellular signalling pathways terminate on transcription factors which in turn activate or deactivate other genes. So for many proteins, their function implies some type of control. But there are also many metabolic genes which function as enzymes in the breakdown of food and as such don't have an implicit control function. But in any case, you're looking at a higher level of organization than I was; you're looking at the purpose of the function and I was just looking at the function itself.
As for your comment about reducing meaning to function, I think it's essential that we realize that there's nothing inherently mysterious or "spooky" about the concept of "meaning". Meaning just implies signal agreement, a correspondence between two things which can send a message. It happens quite naturally in cells all the time, and it just happens that it can also take on abstract forms in humans with complex languages and ideas. I think if we're going to "retrain the scientific imagination to see purpose in nature" we need to start with the purpose and meaning that is concretely instantiated in biological systems, and we can deal with more abstract forms of purpose and meaning later on.
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 20. March 2002 08:17
John:
I am coming at this from the philosophical literature, where those who are naturalistically inclined tend to take a hard-line reductionist position (Daniel Dennett being the most notorious one, but he is pretty characteristic). Other philosophers who are non-reductionist do not typically bother themselves with alternative forms of explanation, being content to just live with the inconsistency (John Searle and Thomas Nagel, for example).
So, maybe my remarks about the reality of the teleological organization of the cell were unnecessary, but they were a reflection of my struggle to have this simple fact recognized by my philosophical colleages, who think the only reality is mechanistic interaction, and everything else is just a "projection" of the human mind, or a Kantian "regulative principle," whatever that is supposed to mean.
I guess we basically agree with each other. But you seemed to be skeptical about what I call my "biofunctional realist" position. I suppose we are talking past each other, somehow. Isn't it the case that most of your scientific colleagues would also say that the functionality of the cell can ultimately be "reduced" to mechanistic interactions? That is what I am disputing.
IP: Logged
|
|
mturner
Member
Member # 190
|
posted 20. March 2002 12:24
**
Hello people,
I know I'm not at your level of discourse, but this topic is very important to me, in my own amateurish way, and so I hope you will indulge me the occasional comment.
My first would be to John Bracht, who said,
"As for your comment about reducing meaning to function, I think it's essential that we realize that there's nothing inherently mysterious or "spooky" about the concept of "meaning". Meaning just implies signal agreement, a correspondence between two things which can send a message."
I do not think of 'meaning' as 'spooky', but it is not 'mechanical',either. Meaning is not mere 'signal agreement'. There is always the semiotic element of 'signal interpretation'. Interpretation is dynamic, not mechanical. Therefore 'meaning' can only be understood from a dynamic perspective. In fact, I would go so far as to say that the entire, "ID versus the Standard Evolutionary Model", issue is founded in the essential ontological positions of Dynamism contra Mechanism. And the root of that difference is that Dynamism allows for 'Intelligence' as a natural force, but Mechanism regards it as nothing but an epiphenomenon of mass/energy.
pax,
mturner
IP: Logged
|
|
|
Printer-friendly view of this topic