Uncommon Descent Serving The Intelligent Design Community

Mathematically Defining Functional Information In Biology

Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

Lecture by Kirk Durston,  Biophysics PhD candidate, University of Guelph

[youtube XWi9TMwPthE nolink]

Click here to read the Szostak paper referred to in the video.

 HT to UD subscriber bornagain77 for the video and the link to the paper.

Comments
Mark Frank No, I was correct in using 'reciprocal'. If the probability of NID is 10:1 the reciprocal is 1:10 Before correcting someone it's advisable to make doubly sure a mistake was made in the first place.DaveScot
January 31, 2009
January
01
Jan
31
31
2009
11:46 PM
11
11
46
PM
PDT
Prof O Yes, as I stated, we can make a probability distribution. According to the laws of physics there could be any number of elementary particles in a universe ranging from zero to infinity. There is no law that prefers any particular number or range of numbers. The current estimated number of solutions to 'string theory' is 10^500. Each solution is a universe with unique parameters. Among that set no one has yet discovered a solution that defines a feasible universe to say nothing of one in which stars and galaxies can form. Unless and until a law is discovered that limits a universe to some range of particle numbers we have a very, very low probability of having just the right number for us to be here talking about it.DaveScot
January 31, 2009
January
01
Jan
31
31
2009
11:30 PM
11
11
30
PM
PDT
Re #163 Dave You ask for a probability of design for a certain thing P(e/ID) Do you mean the probability of design given the result - which is P(ID|e) - or do you mean P(e|ID)? I am guessing you mean P(e|ID) but this confusion runs through your comment. For the same certain thing can you provide a probability for non-design P(e/NID) ?. Well Kirk certainly made an estimate! But see my general comment below. Neither of them are 1, by the way. Agreed If you can give me the probability for your non-design theory then just take the reciprocal and you have the probability for mine. :-p I doubt it. I guess you mean the complement (the reciprocal of any probability less than 1 is greater than 1). But even then P(e|X) and P(e|-X) do not have to add up to 1. If you can’t give me a probability with calculations shown for how you reached it then that puts both theories on equal footing so then explain to me why yours should be taught while mine is excluded? Either way you answer, my point is made. Isn’t that just precious? I am not asking for a precise estimate for P(e|ID) or (e|Darwinism) - which is not the same as NID. I was just asking Kirk to (a) Define what his ID hypothesis actually is (b) Accept that P(e|ID) is less than 1 I actually think the whole probability approach is daft - just too many unknowns and meaningless numbers. But Kirk and much of the ID community make this the foundation of their argument - so I am trying to meet them on their own terms. My reasons for rejecting ID are not to do with probabilities.Mark Frank
January 31, 2009
January
01
Jan
31
31
2009
11:27 PM
11
11
27
PM
PDT
DaveScot[167], Thanks for the info and the link. It is fascinating stuff indeed. Not to belabor the point, but I still wonder how we can do the probability calculation that the filter requires. We must find the probability that the initial number of particles is in the interval 10^80 plus/minus 10^20. But from what probability distribution was this number drawn? Can we really make any resonable assumptions here? What does it even mean that this number was drawn from a probability distribution? Assuming an infinite number of universes would explain everything. Every event, no matter how improbable, occurs infinitely many times. Although I'm on Team Darwin, I wouldn't use it! Getting late, goodnight for real!Prof_P.Olofsson
January 31, 2009
January
01
Jan
31
31
2009
10:48 PM
10
10
48
PM
PDT
Professor O The explanatory filter applied to the fine tuning of the universe. The number of elementary particles in the universe is approximated to be 10^80. There's no known law which says it has to be that number. It could be any larger or smaller number. This establishes the complexity of the number since it one specific one among an infinite or nearly infinite set. Physicists tell us that this specific amount of mass energy, plus or minus about 10^20 particles (about the number of particles in a grain of sand), is required to balance the gravitational force in the universe such that it doesn't prematurely collapse before stars and galaxies could form or fly apart so quickly that stars and galaxies could not form. Thus the number has a specification - it allows stars and galaxies to form. Now that we have established specified complexity we must assess whether or not law & chance could reasonably be responsible for the specified complexity. There is no known physical law that demands any specific mass energy of the universe but that's not to say there is no physical law, just that no one knows of it. There is a chance hypothesis generally called the multiverse which postulates an infinite or nearly infinite number of universes that exist either serially or in parallel and we just happen to be in one which allows stars and galaxies to form. For further reading on this see my article below which describes the current thinking by cosmologists on this issue. Cosmologists and physicists don't like to say it but when pressed they acknowledge that ID is still one explanation on a very short list of things that might explain the fine tuning problem. After 40 years of silence Analog Magazine finally tackles Intelligent DesignDaveScot
January 31, 2009
January
01
Jan
31
31
2009
10:09 PM
10
10
09
PM
PDT
DaveScot[164], Luck or not, you suggested that the filter can be applied and I cannot see how. For the Caputo case and the flagellum, we can compute probailities (at least try to) but here I don't see what we can do.Prof_P.Olofsson
January 31, 2009
January
01
Jan
31
31
2009
09:32 PM
9
09
32
PM
PDT
jerry[162], Thanks. Weinberg is great. In the late 70's I read his book about the "first 3 minutes" and also watched him on TV when he came to Sweden to get the price. I also saw him gave a talk at Rice University, being critical of Thomas Kuhn. Alas he was also critical of the great Swede Hannes Alfven but I suppose I can accept that. Anyways, I still don't understand how one would apply the filter to this fine-tuning problem. It's very different from other problems we have discussed such as the Caputo case or the bacterial flagellum. I can't see how we could possibly compute a probability. It's good that there is accuracy to the 120th decimal place though, otherwise we wouldn't be here to wonder why it is so.Prof_P.Olofsson
January 31, 2009
January
01
Jan
31
31
2009
09:27 PM
9
09
27
PM
PDT
Jerry You're correct. That's the fine tuning problem. If the universe began with an amount of mass energy different that what it was by a single grain of sand then it would have either collapsed under its own weight without forming stars & galaxies or it would have inflated too fast for stars & galaxies to form. There's enough mass in the universe for approximately 10^60 grains of sand. If one more or less we wouldn't be here to number them. What luck!DaveScot
January 31, 2009
January
01
Jan
31
31
2009
09:26 PM
9
09
26
PM
PDT
Mark Frank You ask for a probability of design for a certain thing P(e/ID) For the same certain thing can you provide a probability for non-design P(e/NID) ? Neither of them are 1, by the way. If you can give me the probability for your non-design theory then just take the reciprocal and you have the probability for mine. :-p If you can't give me a probability with calculations shown for how you reached it then that puts both theories on equal footing so then explain to me why yours should be taught while mine is excluded? Either way you answer, my point is made. Isn't that just precious? DaveScot
January 31, 2009
January
01
Jan
31
31
2009
09:17 PM
9
09
17
PM
PDT
Prof_P.Olofsson, This is supposedly from a Scientific American article that is not online so I cannot verify it. It is from a Nobel Laureate named Steven Weinberg. One constant does seem to require an incredible fine-tuning -- The existence of life of any kind seems to require a cancellation between different contributions to the vacuum energy, accurate to about 120 decimal places. This means that if the energies of the Big Bang were, in arbitrary units, not: 100000000000000000000000000000000000000000000000000 000000000000000000000000000000000000000000000000000 000000000000000000, but instead: 100000000000000000000000000000000000000000000000000 000000000000000000000000000000000000000000000000000 000000000000000001, there would be no life of any sort in the entire universe because as Weinberg states: the universe either would go through a complete cycle of expansion and contraction before life could arise, or would expand so rapidly that no galaxies or stars could form. Maybe someone else can verify the accuracy of this and whether it is still current or if Dave was referring to something else.jerry
January 31, 2009
January
01
Jan
31
31
2009
09:09 PM
9
09
09
PM
PDT
Just to clarify my previous comment which I wrote quickly while at a friend's house. Functional complex specified information is 1) in all living things. 2) is in human activity every day. Any written activity counts as functional complex specified activity. Speech is also an example. Now for animals which make signs and communicate there is a similar pattern and the main question I see is if it is complex enough. That could take a whole thread I am sure. 3) is not found anywhere in nature outside of life. 4) human intelligence does not yet have the ability to create life from scratch but has the ability to created additional functional complex specified information within a genome primarily using other proteins and DNA patterns as a template. Now as best as I know none of this additional FSCI created by humans is new but are copies of other sequences or similar sequences. I do not know if anyone has created a new functional protein from scratch. Maybe someone like Kirk would know the answer to that. These are all observations. So if one wants to assign probabilities to anything then these are some of the facts that one could use. Another is 1) the size of the functional complexity of the proteins specified by the DNA for even the simplest cell or 2) the size of the functional complexity for any protein not in these simplest cells and for which no naturalistic path is likely from those proteins in the simplest cell. Maybe before we are finished here we can learn more about assigning probabilities to various proteins that make sense. Anyway I am looking forward to learning more about this sans mathematical notation. Just in case people want to say that they do not know what use functional complex specified information is then here is a brief explanation. It is information that is complex such as DNA that specifies something else such as RNA or a protein and this RNA or protein has a function. Another example is the alphabet and words which form the average sentence that is complex information that communicates (the specification) some thing, action or quality and this communication of a series of inter-related concepts has a function to inform others. Another is a computer code that is also complex information that specifies a series of actions by the computer that has a function such as printing. The first example is in all life and the other two examples are of human activity. As stated above this phenomena exists no where else in nature and this comment itself is an example of functional complex specified information. Now since no one can never say never there might be a time sometime in the future where someone demonstrates that nature can produce functional complex specified information. However till that time the most likely answer is that nature cannot do it. It is not clear that nature has ever added additional functional complex specified information to already existing FCSI. This last proposition might make an interesting discussion. I do not think anything I have said here is circular.jerry
January 31, 2009
January
01
Jan
31
31
2009
08:54 PM
8
08
54
PM
PDT
Jerry @154
You’re assuming your conclusion here. What we observe is human intelligence generating CSI and CSI ostensibly in biological organisms.
I have asserted nothing. I am making observations.
I beg to differ. In 141 you said:
The fact that no one has ever given an example of this happening anywhere else except for human activity establishes the fact that it is unique and that nature does not have the power to do it.
Your conclusion that "nature does not have the power to do it", where "it" is the generation of CSI, does not follow from the observations. It is, in fact, the issue under discussion. MET mechanisms have been shown to be capable of transmitting information from the environment to the population of genomes that arises from the population subject to that environment. It is not obvious where the limits of these mechanisms lie. When we can identify those limits, we can then make the claim you do. Until then, it is one of the ID hypotheses, not by any means a proven fact. JJJayM
January 31, 2009
January
01
Jan
31
31
2009
08:24 PM
8
08
24
PM
PDT
I'd say that intelligent design can be established without a priori understanding of intelligence. The logic is a simple NOT conditional, what can't be produced by one method must be produced by another. If no knowledge of intelligence exists other then chance and law then the other side of the coin logically fits with NOT itself. We can attribute the other side of the coin to design by intelligence, or simply design by x unknown feature. In this case (in the case of KD, WD etc...) we do have knowledge and understanding of what Intelligence can produce and we do have knowledge what chance and law by itself can produce. His conclusions are thus very logical as all ID proponents have been repeating over and over again.ab
January 31, 2009
January
01
Jan
31
31
2009
08:22 PM
8
08
22
PM
PDT
StephenB "Once again, I will go for broke and claim with “virtual certainty” that in no case did law and chance cause the formation." However, since we were created by law and chance, and we created the sand castles, are they therefore not created by law and chance? :) I know my conclusion is resting on an unproven first statement, but I thought I would have a little fun and maybe learn something trying to think like a Darwinist.Peter
January 31, 2009
January
01
Jan
31
31
2009
08:10 PM
8
08
10
PM
PDT
----Rob: "The assumptions that that human design activity does not reduce to law+chance, that specified complexity is a coherent concept, and that humans create it and nature does not are not established facts in science. It seems that establishing those facts, which would involve some empirical work, would be a good first step for ID proponents that want to base their arguments on them." Let’s take two quick examples and work our way down from there: I begin with the one-hundred fifty-five written posts on this thread. Most of them contain over 1000 information bits, and each one is specified. Now I am going to go out on a limb and assert that each instance was generated by human activity. Further, I will argue that not one instance occurred as a result of law and chance. Are you prepared to take up the other side of that argument? It would seem that you are. Would you care to make your case and explain why I have no right to make my claim? Here is another one just for fun: I point now to the numerous sand castles, (2,000,000 grains per construction formed to specificity) that have been observed on the oceans’ shorelines. Once again, I will go for broke and claim with "virtual certainty" that in no case did law and chance cause the formation. Are you ready to argue against that proposition? Will you seriously contend that the formulation is not "coherent."StephenB
January 31, 2009
January
01
Jan
31
31
2009
07:38 PM
7
07
38
PM
PDT
By the way, while we're waiting for Kirk, does anybody know to what DaveScot is referring in his post [119]? How does one apply the filter to the fine-tuning of the universe?Prof_P.Olofsson
January 31, 2009
January
01
Jan
31
31
2009
07:33 PM
7
07
33
PM
PDT
Adel[152], You are very welcome! I hope Mr. Durstan can catch up on his work and his sleep, and be back here to continue the discussion. Perhaps we can sort out a few things regarding the inference and then we'll all disagree in the end! Goodnight y'all.Prof_P.Olofsson
January 31, 2009
January
01
Jan
31
31
2009
07:01 PM
7
07
01
PM
PDT
"You’re assuming your conclusion here. What we observe is human intelligence generating CSI and CSI ostensibly in biological organisms." I have asserted nothing. I am making observations.jerry
January 31, 2009
January
01
Jan
31
31
2009
02:48 PM
2
02
48
PM
PDT
KD Thank you for perservering though these questions. Them must seem very elementary to you considering how much you simplify your arguments and repeat yourself. And yet they still don't get it, or refuse to get it. I have recently read (signficant portions) of Yockey's book "Information theory and molecular biology," a great book. He had already convinced me that it was statistically extremely unlikely for even a small protein to be created by natural methods, so forget about the first life form. So my question to you is, given the improbability of the simplest life being created by natural means, what is the likelihood that all the life forms that emerged in the Cambrian explosion were created from natural processes, given of course that life had already existed. If I understand you correctly then it would be significantly less then 1/10^80,000 since you have already included all trials for the potential natural cause and in this case the amount of functional information required is infinitely greater.Peter
January 31, 2009
January
01
Jan
31
31
2009
02:19 PM
2
02
19
PM
PDT
Professor Olofsson, Thank you, too!Adel DiBagno
January 31, 2009
January
01
Jan
31
31
2009
02:15 PM
2
02
15
PM
PDT
KD, thank you for your great input! It's much appreciated! :Dskynetx
January 31, 2009
January
01
Jan
31
31
2009
02:02 PM
2
02
02
PM
PDT
KD[135], Mr. Dyston, let me join Mark in thanking you for taking the time to explain your statements and results. Let me respond to the posts in which my remarks were addressed. e) Your proposed "intelligence hypothesis" is
An attribute that distinguishes intelligence from mindless natural processes is the ability of intelligence to produce effects requiring significant levels of functional information.
This sounds more like a definition of intelligence than a hypothesis. A hypothesis is a statement that can be true or false, and to which we can assign probabilities. Your hypothesis can only be deemed true or false if we have a separate definition of intelligence. At any rate, it sounds as though your hypothesis really is "ID was required." f) Regarding your claim "It was about 10^80000 times more probable that ID was required," you say
I do not agree that the statement is inaccurately formulated. This is not a probability statement about ID; it is a probability statement about whether ID is likely to be required.
So the intelligence hypothesis is "ID is required." To avoid semantics, let me label it hypothesis "A." In the video you are making a conditional probability statement about A of the type P(A|E) where E represents any evidence under consideration, be it empirical data or something else. However, there is no such calculation to be found in your latest post. Your probabilities are all conditional in the other direction. Same issue further down:
It is not the probability of ID that I am dealing with, but the probability that ID is required.
that is, the probability of A. Yet, nowhere do you compute the probability of A. With your definition of "ID," I don't see the relevance of noticing that P(ID)=1 and P(e|ID)=1. when you really should deal with probabilities P(A), P(E|A), and P(A|E). [Note: I use "E" here to denote generic "evidence" as I don't want to misuse your more specific "e."] Stating that P(ID)=1 merely establishes that we all agree that ingelligence exists. As an event with probability 1 is statistically independent of any other event, what is the point of P(e|ID)? It is equal to the unconditional probability P(e) which is already 1, as "e" demonstrably exists. [Parenthesis: Your bird analogy only illustrates that we can conclude that a set is non-empty by observing one of its elements. It doesn't let us draw conclusions about the individual elements such as "every black bird likes acorns.] Further down you say
...how much more probable is it that ID was required than nature? The answer is P(33 Kbytes|ID)/P(33 Kbytes/B) = 10^80,000 In other words, ID is 10^80,000 more likely to be required...
so you are again making statements about the probability of A: "ID was required," yet, all your probabilities are conditional in the other direction. I am honestly trying to follow your argument. I think my original criticism stands unopposed, but I am certainly willing to listen. I think a first question that must be asked and that I really would like to get answered is: Are you doing Bayesian inference or likelihood comparison? In the video you are making claims that suggests the former but your response above suggests the latter. I understand that my comment will be perceived by many as technical or pedantic, but it isn't. There are huge differences between different types of statistical analyses both in terms of assumpions and conclusions. If you don't believe me, read Dr D's "Elimination vs. Comparison"!Prof_P.Olofsson
January 31, 2009
January
01
Jan
31
31
2009
01:38 PM
1
01
38
PM
PDT
jerry @ 141
You are asking us to establish that 2 + 2 = 4.
I realize that the fundamental claims wrt specified complexity are as obvious as 2+2=4 to you, but after 10+ years of making those claims, the ID community still hasn't gotten any traction with them, at least not as far as mainstream science is concerned. I think that operationalizing and scientifically testing these claims would do a lot in this regard.R0b
January 31, 2009
January
01
Jan
31
31
2009
01:29 PM
1
01
29
PM
PDT
KD @ 135:
If Rob wants to talk about carbon dust, then a very large number of configurations are possible (N=a large number), but almost any one of those piles of dust has the same function (the basic properties of piles of carbon dust), so M(Ex) = a large number and M(Ex) is approximately equal to N, with the result that the amount of functional information required is close to zero.
I'm talking about all possible configurations of the 10^23 atoms of carbon, which includes chunks of crystal, dust, and free-floating atoms. As you said, the number of configurations is large, and that's putting it mildly. If we're defining functionality in terms of hardness, the diamond lattice configuration marks a functionality threshold, below which are all but an infinitesimal fraction of the possible configurations. How does that not constitute a huge amount of functional information?R0b
January 31, 2009
January
01
Jan
31
31
2009
01:09 PM
1
01
09
PM
PDT
jerry @ 102:
Diamonds don’t specify anything. They might be functional in some contexts and so may be a rock if you are defending yourself or build a wall.
I didn't say that diamonds specify anything. And I agree that both diamonds and rocks can be functional. Kirk Durston certainly hasn't disputed the functionality of diamonds. Dave @ 121:
A diamond does not rise to the level of specified complexity. While one might make a case that it has specification, as an abrasive for instance, it is not complex as the atoms are in a simple repetitive arrangement i.e. law and chance is a perfectly valid explanation for the structure of a diamond.
I explicitly said in the comment you referenced that I'm not talking about specified complexity, but rather Durston's functional information. And saying that a diamond is not complex because of its simple repetitive structure is a non sequitur. Many of Dembski's examples of complexity are simple and repetitive. Your "i.e." makes more sense.R0b
January 31, 2009
January
01
Jan
31
31
2009
12:53 PM
12
12
53
PM
PDT
Jerry[140], I didn't ask about the numbers, I asked about the logic. Is Kirk assessing how likely an outcome is under the ID assumption or how likely ID is given the outcome?Prof_P.Olofsson
January 31, 2009
January
01
Jan
31
31
2009
11:50 AM
11
11
50
AM
PDT
JayM: there has been a little discussion here on the subject of calculating CSI, mostly by gpuccio and kairosfocus. The formula gpuccio suggested is in fact Kirk's one we're discussing here, once the specification has been given. (however he seems to equate complexity with improbability, which as Davescot has already shown us is wrong; stuff can be improbable and simple at the same time.)Venus Mousetrap
January 31, 2009
January
01
Jan
31
31
2009
11:30 AM
11
11
30
AM
PDT
KD @ 135 Again thanks for putting so much work into your comments. There is so much that I would like to say in response but this thread is already excessively long. So I am going to concentrate on one issue. What exactly is the "ID" hypothesis for the presence of proteins with high functional information in living things? You say: Intelligence hypothesis: An attribute that distinguishes intelligence from mindless natural processes is the ability of intelligence to produce effects requiring significant levels of functional information. That is a hypothesis (or possibly a definition) about intelligence but it is not a hypothesis about how functional proteins came to exist. What is it that you are proposing lead to the presence functional proteins in living things (it certainly wasn't Craig Venter or any other human). Without a hypothesis to explain the outcome we can't begin to talk about the probability of the outcome given the hypothesis or vice versa.Mark Frank
January 31, 2009
January
01
Jan
31
31
2009
11:14 AM
11
11
14
AM
PDT
Jerry @141
The fact that no one has ever given an example of this happening anywhere else except for human activity establishes the fact that it is unique and that nature does not have the power to do it.
You're assuming your conclusion here. What we observe is human intelligence generating CSI and CSI ostensibly in biological organisms. The ID hypothesis is that that CSI is also of intelligent origin. We need to prove that, not simply assert it. One step toward doing so is to clearly describe how to compute CSI in biological systems, knowing what we do of MET mechanisms. As someone pointed out earlier in this thread, that hasn't yet been done (at least, I haven't been able to find such documentation after considerable searching). JJJayM
January 31, 2009
January
01
Jan
31
31
2009
11:13 AM
11
11
13
AM
PDT
----Professor Olofsson: "There are certainly mathematicians on “our side” who work on applying math to evolutionary biology (such as Rick Durrett at Cornell) but I agree that much more can be done." This is all I have been asking for from "your" side. Provide me with your numbers and allow us to scrutinize them with the same rigor that you scrutinize ours. I say that the Darwinistic paradigm does not even come close to ID in lending itself to mathematical models. You don't seem willing to admit or even comment on the fact that this reciprocal point is cricital to the overall debate. Rather, you seem to exempt yourself from that side of the issue even though your mathematical credentials are sufficient to address it. Are you now saying that the Darwinist formulation is precise enough to be measured mathematically? Isn't it time for you or someone in your camp to provide some semblance of an argument on this subject?StephenB
January 31, 2009
January
01
Jan
31
31
2009
10:50 AM
10
10
50
AM
PDT
1 4 5 6 7 8 11

Leave a Reply