Mathematically Defining Functional Information In Biology

_{January 28, 2009

Intelligent Design}

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 (#241): Just two points. 1) "How would you know that intelligence had not been involved? Whatever series of events you observed they may have been designed." Well, again I cannot agree with you on that kind of argument. You have no faith in the objectivity and serious scientific approach of us IDists! In other word, if some structure (like the flagellum) which, according to ID theory, has CSI and IC, could be formed in the lab under controlled condition, and if we could understand the way it happens, that would falsify the point of ID. The idea that God could simulate an apparently perfect spontaneous and unguided process has nothing to do with ID. That could be an extreme (and rather pitiful) line of defense for a religious position, but ID is not based on any religious position. So, we would very simply accept the evidence. The point of ID is that we really are convinced that it cannot happen that way. If we see it happen that way, and understand how, then we were wrong, period. 2) Just a clarification about the CSI (and IC) in the flagellum. I think there is some confusion about that. First of all, we have a lot of CSI in all the individual proteins which make up the flagellum (I think they are almost 50). The TTSS can be used to make a very partial (and extremely unsuccesful) generic model for some of them, but not for all the others. And anyway, even if someone still believes that the TTSS is in some way a precursor of the flagellum (which I don't), the CSI in the TTSS still remains to be explained. But that's only the CSI at the level of single proteins. Obviously, there is much more in a complex, and IC, machine like the flagellum. There is the huge engineering necessary to design and put together all those individual components so that the global machine can arise and work. That the true IC, and certainly a lot of CSI. But you know my point: that kind of "higher level" CSI is very difficult to model, and that's why I never try to do that, and stick to the CSI of single proteins. Personally, I find that even Dembski's model of the shopping cart is not a very satisfactory approach (on that, I have to agree with Prof. Olofsson). But the fact that something is difficult to model does not in any way mean that it does not exist. Higher level CSI is everywhere in the biological world, in organization, regulation, error checking, body plan control, and so on. And sometime in the future, we will have to find a quantitative way of approaching that. In the meantime, for the flagellum, the "qualitative" approach to IC made by Behe in DBB remains absolutely valid, while for a real quantitative approach to higher level CSI in the whole machine we will probably have to wait. But we can always analyze quantitatively the cumulative CSI present in the single proteins.gpuccio_{February 2, 2009
February
02
Feb
2
02
2009
11:39 AM
11
11
39
AM
PDT}

Professor -- Distilled argument: If P(E|D) is n times as large as P(E|C), you cannot conclude that P(D) is n times as large as P(C ). I'll agree but if n is large enough it should still make you hmmmm. To revisit the Sally Clark case: 2 infants die in her care within a relatively short time. SIDS is diagnosed as the cause. Do you: ignore it and say "no big deal, that sort of thing happens?" or investigate for the sake of future children?tribune7_{February 2, 2009
February
02
Feb
2
02
2009
11:33 AM
11
11
33
AM
PDT}

Peter @252

According to MET, the flagellum arose from similar precursors via one or more of a set of natural mechanisms.
You left out the most important word two words! “According to MET, the flagellum may have arosen from similar precursors via one or more of a set of natural mechanisms.”
I see your point, but that's why I used the phrase "According to MET." In the context of MET, there is no "may have."
Evolution is based on supposition and has not been proven.
That overstates the case. Many examples of microevolution have been observed and documented. Any viable ID theory has to take that into account. All of this is beside the point, however. DaveScot's use of "ex nihilo" suggests that MET does say that the flagellum arose from nothing, which is not at all the claim. JJJayM_{February 2, 2009
February
02
Feb
2
02
2009
11:33 AM
11
11
33
AM
PDT}

Hi Prof. Olofsson, your participation and expertise on the issue is much appreciated. Your honesty is quite the turn around. Because of this I believe you have generated much respect from many ID proponents. I regard such criticism as being healthy for ID in general. I only have a few questions regarding your back-to-back with DaveScot. You say: Nevertheless, it is what the explanatory filter requires. Approximately how many chance hypothesis could there be? Is the other side proposing an infinite chance hypothesis to explain away biological systems as they have already explained away the anthropic principle with multi-verse theory? If so, how would you attempt to plug that into an equation? Isn't this equivalent to saying that 'God-did-it' since if you cannot even narrow down a chance hypothesis to specifics then you infer multiple or infinite sets of non-specifics to explain away the details? When does a chance hypothesis begin and when does one end? (Please note that I am not directly targeting these questions at you)ab_{February 2, 2009
February
02
Feb
2
02
2009
11:31 AM
11
11
31
AM
PDT}

-No, my dear friend, Demsbki does not and I don’t even have to look at your link to know! Professor, I didn't read the book but I'll go on a limb and say I think his reasoning is that a 20 specific amino acids have to organize significantly large chains in a specific manner to form specific proteins which then have to organize in a specific manner to form the flagellum. It's something that can't happen by chance.tribune7_{February 2, 2009
February
02
Feb
2
02
2009
11:24 AM
11
11
24
AM
PDT}

You can’t use significance levels of 10^-150, Sure you can. To illustrate the impossible. Besides, it’s not about the numbers per se, it’s about the logic. Professor, did you understand what I wrote? I was agreeing with you.tribune7_{February 2, 2009
February
02
Feb
2
02
2009
11:16 AM
11
11
16
AM
PDT}

trib, "I thought we might not want to go more off-track with a debate about CSI" Agreed.Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
11:15 AM
11
11
15
AM
PDT}

tribune[250], No, my dear friend, Demsbki does not and I don't even have to look at your link to know! He never considers "design hypotheses" as his approach is strictly eliminative. Nice try but no cigarillo!Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
11:13 AM
11
11
13
AM
PDT}

JayM [245] "According to MET, the flagellum arose from similar precursors via one or more of a set of natural mechanisms." You left out the most important word two words! "According to MET, the flagellum may have arosen from similar precursors via one or more of a set of natural mechanisms." Evolution is based on supposition and has not been proven.Peter_{February 2, 2009
February
02
Feb
2
02
2009
11:11 AM
11
11
11
AM
PDT}

tribune[240], Distilled argument: If P(E|D) is n times as large as P(E|C), you cannot conclude that P(D) is n times as large as P(C ).Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
11:10 AM
11
11
10
AM
PDT}

I still see no probability calculation in your reply. I'll appeal to authority. Dembski refers to it here and apparently goes into more detail in No Free Lunch. only designed objects have CSI and the flagellum structure exhibits CSI . . . .you are, with all due respect and in congeniality, somewhat vague. Well, I did say I was leaving that aside :-) I thought we might not want to go more off-track with a debate about CSI. The point, of course, is the design hypothesis of the flagellum.tribune7_{February 2, 2009
February
02
Feb
2
02
2009
11:09 AM
11
11
09
AM
PDT}

tribune[240], First, the UPB is not relevant in everyday statistical analysis. You can't use significance levels of 10^-150, surely you understand that? Besides, it's not about the numbers per se, it's about the logic. If you don't even want to try to understand what I write, I'll stop. I understand the kneejerk reactions of many here to debate every single word uttered by an ID critic such as I, but actually, none of my posts in this thread are anti-ID.Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
10:34 AM
10
10
34
AM
PDT}

StephenB[243], You don't accept arguments from authority unless, of course, you make those arguments yourself such as in [220], by referring to Durston.Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
10:26 AM
10
10
26
AM
PDT}

tribune[244], I'm not trying to disprove ID or prove evolution, I'm just asking what you'd do. I still see no probability calculation in your reply. When you say
only designed objects have CSI and the flagellum structure exhibits CSII
you are, with all due respect and in congeniality, somewhat vague.Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
10:20 AM
10
10
20
AM
PDT}

DaveScot[231],
You say we can never rule out all chance hypotheses. That is true but, happily, we don’t need to.
Nevertheless, it is what the explanatory filter requires.Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
10:11 AM
10
10
11
AM
PDT}

DaveScot @231
Thus we can state the ID hypothesis for the flagellum as: The ex nihilo creation of a flagellum cannot occur absent intelligent agency.
That's a curious formulation given that modern evolutionary theory (MET) does not suggest that a flagellum can be created ex nihilo either. According to MET, the flagellum arose from similar precursors via one or more of a set of natural mechanisms. JJJayM_{February 2, 2009
February
02
Feb
2
02
2009
09:59 AM
9
09
59
AM
PDT}

But what do we do in practice? Consider the flagellum. Now give me a design hypothesis and tell me what probability it gives the formation of the flagellum. Leaving aside the positive argument that only designed objects have CSI and the flagellum structure exhibits CSI, you have the serendipituous ordering of the proteins, for which the anti-IDist invokes "the-unkown-law-of-the-gaps. and IC which says that Darwinian evolution is incapable of forming it, which the anti-IDists rebut with "can so". Now, I'm not saying that last rebuttal is completely without merit but it has not been definitely made despite the screaming.tribune7_{February 2, 2009
February
02
Feb
2
02
2009
09:48 AM
9
09
48
AM
PDT}

-----Rob: “I’m saying that “specified complexity” has not been accepted as a legitimate scientific concept by the scientific community.” You must be very new here, so I will try not to be unduly harsh. Suffice it to say that we don’t accept arguments from authority on this site, because the so-called “authorities” have long since lost their credibility. Evolutionary biology is a monolithic monster that survives solely by misinforming the public about the current status of evolutionary theory and by attempting to discredit ID scientists even to the point of slander. Do you know anything at all about ID's current history?StephenB_{February 2, 2009
February
02
Feb
2
02
2009
09:46 AM
9
09
46
AM
PDT}

Kirk, I understand perectly well if you choose to stay away altogether; there's a lot going on here and responding to everything will take a lot of time away from you. I asked above if it's possible to create a new thread that could stay focused, at least for a while. A think my question for you whether you wish to do a Bayesian analysis or a comparison of likelihoods is a good starting point. Regardless of what the biology is or how you define information etc, in the end you need to decide how to use your probabilities and what type of inference to draw. My main point is that your statement in the video is inherently Bayesian, yet, your explanation above is not.Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
09:33 AM
9
09
33
AM
PDT}

the ID hypothesis for the flagellum as: The ex nihilo creation of a flagellum cannot occur absent intelligent agency. The hypothesis can be falsified by a single observation of a flagellum forming ex nihilo by law and chance alone How would you know that intelligence had not been involved? Whatever series of events you observed they may have been designed.Mark Frank_{February 2, 2009
February
02
Feb
2
02
2009
09:27 AM
9
09
27
AM
PDT}

Professor, A couple points with regard to guilt, statistics and the case you cite: One in a million does not even come close to approaching the UPB. Many of us might know someone who has had a one-in-a-million experience. If you had a one-in-a-million experience you would feel yourself quite fortune (or unfortunate) but you would recognize they do happen. Now, if you experienced something with a probability of 10^150 . . . There are, however, more significant reasons why we can't be certain of Ms. Clark's guilt -- which apparently based solely on the probability of SIDS -- to the point of conviction. Both children not dying of SIDS -- and they probably didn't -- does not mean Ms. Clark murdered them. One or both could have died of something other than SIDS or intent by Ms. Clark and the probability calculations would still hold true. Misdiagnosis by law enforcement or medical personnel is certainly within the realm of possibility. And of course, -- to be hard-nose ugly for a second -- just because a direct connection to murder can't be made doesn't mean it didn't happen.tribune7_{February 2, 2009
February
02
Feb
2
02
2009
09:04 AM
9
09
04
AM
PDT}

gpuccio:
I think you are confounding facts with assumptions. In science, one has to start the discussion with “facts” that are already established, and then anyone can make any reasonable assumption about those facts.
I should say "premises" rather than "assumptions". And yes, scientific arguments should be premised on established facts.
As for me, I do prefer not to “be a part of mainstream science”
Ah, then my comments don't apply to you. I had assumed that ID proponents in general were bothered that ID is "expelled" from mainstream science. Maybe I'm completely wrong about that.
Your suggestion, that we should “start” with the “established assumption” that “human design activity reduces to law+chance” (it is, indeed, the preferred assumption of most scientists today), just because it is the prevailing assumption in a social circle, is called scientific tyranny.
Where did I suggest that?
That specified complexity is a coherent concept: that is not an assumption:
A premise, then.
Now, that is a very simple and unequivocal definition of CSI. Wee can discuss about what we mean for funtion, or about how the complexity can be calculated in specific cases, but that does not make the definition incoherent.
But it does make it equivocal, and we can't establish its coherence until we disambiguate it.
just take this (rather long) post. It is CSI in that sense, without any doubt. Are you doubting that?
It depends on how we disambiguate your definition of CSI. If the improbability is conditioned on all law+chance hypotheses, the we both have reason to question whether these posts contain CSI. You said yourself that we don't know for certain if human design activity reduces to law+chance.
I am not making a logical statement here. I am just making an objective statement about what we have so far observed (an empirical statement). If my statement is wrong, then you can easily show why: just show us a known example of spontaneous CSI in nature.
Again, it depends on how we disambiguate CSI. It may be that nature can't produce CSI by definition, in which case the statement isn't empirical but tautological. And I'm not saying that your statement is wrong, only that the ID community has provided no data to support it. You could, for instance, run a double-blind test where a large number of different subjects independently calculate the CSI of objects whose origins they don't know. Maybe throw in some large natural crystals (see the Cueva de los Cristales), or some accidentally discovered complex abstraction, like the Mandelbrot set. The point is that if you want your premises to be accepted as science, then you have to do the science.R0b_{February 2, 2009
February
02
Feb
2
02
2009
08:58 AM
8
08
58
AM
PDT}

It seems that UD would be wise to create a parallel forum to allow all of the various threads and sideshows to flourish.William Wallace_{February 2, 2009
February
02
Feb
2
02
2009
08:52 AM
8
08
52
AM
PDT}

I am very tempted to respond to some of the points that have been raised here, but I can see that there is a major problem with this discussion in that there are far too many ideas being discussed in too loose a fashion. I also see that some of the problems in this discussion are resulting from the convergence (or joint application) of logical argumentation, empirical probabilities and Bayesian probabilities. I would like to start a new thread that works through my thinking on this subject, but proceed in a meticulous (pedantic) fashion. I must say in advance, however, that I can only devote a small amount of time per week on this discussion, so it would proceed at a slow, but hopefully quality, pace. Still, I feel that a meticulous, slow-paced discussion can be much more productive than a fast-paced, multi-issue, loose one. I don't know how to start a new thread on this forum, but if this is agreeable to the powers that be, then those powers should let me know how to start a new thread and I'll proceed. I would make the first post, lay out a couple ground rules to keep the discussion tightly focused, present a couple definitions, and then pause to make sure we are in sufficient agreement to proceed. Once there is sufficient agreement on the basics, we can then proceed to one or two subsequent points, discus them, then move to the next point or two, and so on. In that way, I think we can accomplish something. We may not arrive at checkmate, but I am certain that we can at the very least make much better progress toward that end.KD_{February 2, 2009
February
02
Feb
2
02
2009
08:51 AM
8
08
51
AM
PDT}

tribune[229], Good. Of course we can't. What we need is the probability of guilt conditioned on the evidence, hence we need a prior probability of guilt to apply Bayes rule. We can estimate that one from data by looking at the incidence of double infanticides. Whatever it is, it is small. If it is also 1 in a million, the probability of guilt given the evidence becomes 1/2 which makes perfect sense.Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
07:30 AM
7
07
30
AM
PDT}

kairosfocus[228], It's more illustrious to be number one than number two, even if counted from the bottom. I bet the Bhutanis regret scoring all those goals!Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
07:26 AM
7
07
26
AM
PDT}

This thread is getting very long and diverse. As I would like to hear what Kirk has to say and focus the discussion on his use of probabilitic analysis, is there any chance we can create a new thread?Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
07:08 AM
7
07
08
AM
PDT}

tribune[216],
What would convince you that it’s about design and not the designer?
If you say so, that's enough for me. But what do we do in practice? Consider the flagellum. Now give me a design hypothesis and tell me what probability it gives the formation of the flagellum.Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
07:06 AM
7
07
06
AM
PDT}

DaveScot[231], Your reasoning is all in the context of elimination and falsification which is fine, but what we have mainly discussed in this thread is likelihood comparison and Bayeisan inference. In that context, your flagellum hypothesis has no meaning because it does not confer a likelihood on the data under a separate design hypothesis about the existence of design. Your hypothesis immediately confers probability 0 under the chance hypothesis and you cannot assign to it a prior. As for fine-tuning, my questions was what probability distribution you propose. In #170 you said it can be made. Nevertheless, thanks for you posts. They are very informative and interesting.Prof_P.Olofsson_{February 2, 2009
February
02
Feb
2
02
2009
07:01 AM
7
07
01
AM
PDT}

To DaveScot: "The ex nihilo creation of a flagellum cannot occur absent intelligent agency." With what other theory are you comparing ID to that can hypothesise a flagellum appearing out of nowhere? Evolution says the flagellum is built slowly and stepwise not out of nothing, our (ID) problem with that is that we think some of the steps were impossible.GSV_{February 2, 2009
February
02
Feb
2
02
2009
06:47 AM
6
06
47
AM
PDT}

Prev 1 2 3 4 5 … 11 Next

You must be logged in to post a comment.

Leave a Reply