Uncommon Descent Serving The Intelligent Design Community

Of Pulsars and Pauses

Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

DrREC is not just any Darwinist.  He holds a doctorate and has published on complex matters of biology in peer reviewed journals.  He is not stupid.  That’s why I like to use his examples in my posts.  I am not picking on a defenseless layman.  He’s among the Darwinists’ best and brightest.  So let’s get to his latest pronouncement from on high:

DrREC writes: 

Pulsars often have a complex behavior. But is it specified? If we took the pattern of pulses we detect as the ‘design specification’ — the pattern we search for, we would conclude yes. Totally and undeniably circular. Prove me wrong.

Here’s the problem with DrREC’s reasoning.  He seems to assume (despite being told the contrary numerous times), that any “pattern” can be designated post hoc as “specified.”  He does not seem to understand the most basic concepts of design theory.  The answer is that not any pattern can legitimately be called a specification. 

In a comment to my prior post Bruce David explains the concept nicely as follows:

Dembski’s work builds on that of earlier probability theorists’ who were wrestling with the problem that, for example, any pattern of heads and tails obtained by tossing a coin 100 times is equally improbable, yet intuitively, a pattern of 50 heads followed by 50 tails is in some sense far less probable than a ‘normal’ random pattern. In order to solve this conundrum, they came up with the idea of specification—if the pattern of heads and tails can be described independently of the actual pattern itself, then it is specified, and specified patterns can be said to be non-random. And note, the pattern does not have to be described ahead of time; the requirement is just that it is capable of being described independently of the actual pattern itself. In other words, a normal ‘random’ pattern can only be described by something equivalent to ‘the first toss was heads, the second heads, the third tails,’ and so on, whereas the example above is specified because it can be described as I already have, namely, ’50 heads followed by 50 tails’.

Back to DrREC’s question.  The pulses from the pulsar are indeed highly complex (i.e., improbable).  But they are never specified because they cannot be, as Bruce says, “described independently of the actual pattern itself.”  Therefore if we “took the pattern of pulses we detect as the ‘design specification'” even though that pattern could not be described independently of the actual pattern itself, we would simply be wrong.  That pattern does not conform to the definition of a specification. 

DrREC basically says, “If we call any pattern we find a “specification” then any pattern we find will be a “specification,” and that gets us nowhere.  Well, of course he is right as far as it goes.  But at a deeper and more meaningful level he is wrong, because no one says you can call just any pattern you find a specification.  The pattern must conform to a strict criterion before it can be considered a specification. 

So DrREC, I answered your question.  While we are on the issue of pulses you can answer mine.  Suppose researchers detect a repeating series of 1,126 pulses and pauses of unknown origin.  The pulses and pauses start like this (with one’s conforming to pulses and zero’s conforming to pauses):  110111011111011111110 . . .  After analyzing the series they determine that the zero’s are spaces between numbers and the one’s add up to numbers.  Thus, the excerpt I reproduced would be 2, 3, 5 and 7, the first four prime numbers.  The researchers suddenly realize that the 1,126 pulses and pauses represent the prime numbers between 1 and 100.  (Obviously, this was the series in the movie Contact).

My question for you DrREC is this:  Would you join Arch-atheist, uber-materialist, Darwinist Carl Sagan and conclude that this series is obviously designed by an intelligent agent?  If so, why?  After all, it is a hard fact that this series of 1,126 pulses and pauses is NO MORE IMPROBABLE than any other series of 1,126 pulses and pauses.

Comments
It's true that it is possible to translate a statement into gibberish. That is what's done in encryption. But that's not my scenario. I'm asking if you can distinguish an unencrypted sequence with one character out of place from a random sequence. I do not see functional DNA sequences as random. My theory is that they have been built incrementally. What's the alternative? If you cannot build them incrementally, how do you go about finding functional sequences?Petrushka
December 15, 2011
December
12
Dec
15
15
2011
01:47 PM
1
01
47
PM
PDT
...all this takes for granted (assuming the speculation) is the system of coordinated physical protocols which allows the transfer of information in the first place. Do you have any examples of the onset of those?Upright BiPed
December 15, 2011
December
12
Dec
15
15
2011
01:47 PM
1
01
47
PM
PDT
9.3.1.1.1 kairosfocus "Really! Just now, can you report to us on how you composed your rebuttal post by chance keystrokes that got lucky . . .? Case over." Is there any doubt I designed my post? I see you aren't intelligently defending your reduced chi metric.DrREC
December 15, 2011
December
12
Dec
15
15
2011
01:41 PM
1
01
41
PM
PDT
"SHOW us complex protein origination or similar origin of functionally specific strings by blind processes, then tell us." "Cross-species analysis revealed interesting evolutionary paths of how this gene had originated from noncoding DNA sequences: insertion of repeat elements especially Alu contributed to the formation of the first coding exon and six standard splice junctions on the branch leading to humans and chimpanzees, and two subsequent substitutions in the human lineage escaped two stop codons and created an open reading frame of 194 amino acids. We experimentally verified FLJ33706's mRNA and protein expression in the brain." http://www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fjournal.pcbi.1000734DrREC
December 15, 2011
December
12
Dec
15
15
2011
01:40 PM
1
01
40
PM
PDT
Dr Rec: If you want us to accept the equivalent of a perpetual motion machine, SHOW us complex protein origination or similar origin of functionally specific strings by blind processes, then tell us. All you are telling me absent such, is that you are under the control of an a priori that blocks you from acknowledging how hard it is to get to functionally specific sequences by blind processes. If you saw rocks tumbling down a hillside, would you be surprised to see them say falling out in the pattern: Cicero was right? Why or why not? KFkairosfocus
December 15, 2011
December
12
Dec
15
15
2011
01:31 PM
1
01
31
PM
PDT
P, You sound like someone who has never had to design or implement a complex object that had to get mutually fitting parts to work. Recall, that 3 of 64 possible codons mean STOP. Then, tell us about how chance variations of codons can write complex proteins of novel function and regulate their production and transport to the right site where lo and behold all is in place to wait for them. Then, recall also that the proportion of proper fold domains is like 1 in 10^70 of the AA sequence space. The algorithmic challenge to get to functional sequences is real, and tightly constraining. Just as are the challenges to get to a viable post in English by blind processed. These are obvious to all, save those who have long since swallowed the materialist a prioris that t6ell them what MUST have been so. As I noted above, Cicero nailed it c 50 BC. Next time you want to promote the equivalent of a perpetual motion machine, my answer is the same: SHOW and then tell. GEM of TKIkairosfocus
December 15, 2011
December
12
Dec
15
15
2011
01:24 PM
1
01
24
PM
PDT
Welcome back CJYmanUpright BiPed
December 15, 2011
December
12
Dec
15
15
2011
01:22 PM
1
01
22
PM
PDT
Really! Just now, can you report to us on how you composed your rebuttal post by chance keystrokes that got lucky . . .? Case over.kairosfocus
December 15, 2011
December
12
Dec
15
15
2011
01:16 PM
1
01
16
PM
PDT
Then you can answer my question about how to distinguish a sequence that is one base pair from being functional, from a collection of randomly generated sequences
The resultant product of feeding the sequence into the processing equipment (which embodies the protocol) is what distinguishes a random string from a functional string. If it’s only one base pair off, I would imagine that the output would still be considered functional, if tolerances allow.M. Holcumbrink
December 15, 2011
December
12
Dec
15
15
2011
01:07 PM
1
01
07
PM
PDT
We do? Could you give an example of each in biology?
Alternative splicing is algorithmic compression, e.g. splicing together portions of code from various genes to make a new sequence of code. Regulated stepwise procedures (algorithms) are required make this happen. And any time you have a gene that makes protein A, then you jump over one base pair to code for protein B, you have multilayered encryption. Do you deny that this happens in the cell?M. Holcumbrink
December 15, 2011
December
12
Dec
15
15
2011
12:50 PM
12
12
50
PM
PDT
Petrushka,
It’s your problem, not mine.
No, really, it's your problem, because you said
If you cannot distinguish a degraded functional sequence from a random sequence, you cannot assign bits of information to it.
To test whether you actually believe this or are applying it selectively, I posed a simple scenario:
If you give instructions to a man and he translates them in language you don’t understand to a second man who then follows most but gets a few things wrong, you cannot ascribe any functional information to those translated instructions because you don’t know which part was degraded.
In this scenario, because you do not speak the language you cannot tell which information is degraded nor can you distinguish any of it from random gibberish. It seems to test your assertion perfectly. You "you cannot distinguish a degraded functional sequence from a random sequence." Your assertion is simple, and so is my illustration, which demonstrates that what you assert is false.ScottAndrews2
December 15, 2011
December
12
Dec
15
15
2011
12:33 PM
12
12
33
PM
PDT
I have a six-foot wooden plank. I’ve demonstrated repeatedly that I can use it to bridge seemingly impassible gaps. I have video.
You have no theory to back up your characterization of functional space as unbridgeable. Comparative genomics indicates that sequences of cousin species have differences that form a nested hierarchy. When actual gaps have been tested, there are viable intermediate sequences.Petrushka
December 15, 2011
December
12
Dec
15
15
2011
12:29 PM
12
12
29
PM
PDT
There is similar work that gives much lower values, like 1 in 10^24 (Taylor et al., PNAS 98, 10596-10601, 2001). Big number, but with 10^30 or so bacteria on earth, with 1000's of genes, times many generations..... Axe's work doesn't explore all of sequence space, or take into account other families with similar activities. As Petrushka points out, it also doesn't show the functional sequences are isolated or impossible to reach by evolution.DrREC
December 15, 2011
December
12
Dec
15
15
2011
12:23 PM
12
12
23
PM
PDT
If you give instructions to a man and he translates them in language you don’t understand to a second man who then follows most but gets a few things wrong, you cannot ascribe any functional information to those translated instructions because you don’t know which part was degraded.
It's your problem, not mine. My argument is that functional sequences are what they are because they have been built incrementally. I recognize that this requires a sequence space that supports bridges and ridges. That's why the work of and Thornton are important. If there is no way to improve sequences incrementally, evolution is impossible. ID seems to argue that there is a shortcut to evolution, that one can somehow design a long functional sequence without building it incrementally. If so, one should be able to distinguish a sequence that has 499 out of 500 words in place, without testing it in a living system. It seems to me, however, that ID has no theory of functional sequences or how to recognize them or design without doing the chemistry. No theory that specifies the shortest possible functional sequence and no theory of how to design them without starting with known functional sequences. Gpuccio's calculation involves a step function. A sequence is either functional or not functional. If it's functional he declares that all the bits are necessary, but he has no theory of why this should be the case. He has no theoretical reason why a sequence could not be the current state of a series of incremental changes. To make this kind of assertion, one would need to be able to demonstrate that a sequence X, that is one character short of functional, has X - 1 characters specified.Petrushka
December 15, 2011
December
12
Dec
15
15
2011
12:22 PM
12
12
22
PM
PDT
I have a six-foot wooden plank. I've demonstrated repeatedly that I can use it to bridge seemingly impassible gaps. I have video. I've also used it to walk to Hawaii and back. Don't laugh. I told you it serves as a bridge between points. Don't ask me to show you, either. Just go with the extrapolation.ScottAndrews2
December 15, 2011
December
12
Dec
15
15
2011
12:19 PM
12
12
19
PM
PDT
"A protein family is observed" And you go out into the archery field, and draw a bulls-eye around it-as though that were an intended target. Further, you define what belongs in that family, and what doesn't. "its variability while retaining function is used to quantify the info in the AA sequence" No response whatsoever to the flaws in that technique. "Function" is also defined by you. The only function evolution cares about is an increase in fitness. "a functional family" What about all other families that could perform that function, and all potential families? If you really believe in this method, you should at least be able to answer this from above: "The number of fits for some whole domains and enzymes is well below the universal probability bound. If two of these recombine (which is almost a given in some organisms) the number would be pushed over the universal probability bound. So apparently natural+natural undergoing a natural (and highly probable) process yields design detection."DrREC
December 15, 2011
December
12
Dec
15
15
2011
12:15 PM
12
12
15
PM
PDT
I’m still waiting for someone to explain how they would detect design without a pre-established or independently known pattern.
How about the storage of symbolic information arranged in sequences that, when translated using a specific protocol, produces components that interact with each other as if were known in advance, when the information was first stored, what the output would be if they were translated using that protocol? I'm pretty sure that's a pattern. I'm referring, not to the arrangements of DNA, but to the behavior of the entire system. Does anything else do that? Are there any manufacturing processes that start with an abstract representation of components, that, when manufactured, function together? It's quite simple. Detecting design does not involve matching the finished product to a previous specification, or arbitrarily determining that the finished product was specified. It is the presence and application of abstract information itself that is a pattern.ScottAndrews2
December 15, 2011
December
12
Dec
15
15
2011
12:11 PM
12
12
11
PM
PDT
The problem with the "very narrow" argument is that when it has been tested, as in the work of Thornton, there are bridges between variant versions of sequences in living organisms. It doesn't matter how sparse functional space is if it is bridgeable.Petrushka
December 15, 2011
December
12
Dec
15
15
2011
12:02 PM
12
12
02
PM
PDT
This ignores: I) Evolution has no target, where this metric narrowly specifies one II) Other insulin and insulin like sequences that perform the same function have been lopped out III) This isn’t a true mapping of functional space-other proteins not yet sequences, or that don’t even exist in nature that could substitute
This is where the work of D. Axe & others becomes important. What exactly is the best estimate of this functional space? Based on their work, best as I can tell, this functional space is *very* narrow. With that said, I find this to be analogous to the macro world of mechanical components. These components as a necessity must be defined with tolerances. There are therefore ranges of sizes, profiles, orientations and form for particular features that are functional (i.e. a functional space). So for less critical features, the tolerances are relatively loose (larger functional space), but for the business end of these components, the tolerances need to be relatively tight (smaller functional space). And if these tolerances can accommodate certain components being utilized in other systems, with only very minor adjustments, then it is the engineer’s duty to factor that in as well (which is certainly the case for engineered components at the macro level). And best as I can tell, this is exactly what we see in molecular biological components (e.g. horizontal gene transfer). Maybe “brainwashed” and “incorrigible” was a little harsh. But that’s sure what it looks like. You are at least stuck in a rut, it seems to me.M. Holcumbrink
December 15, 2011
December
12
Dec
15
15
2011
11:52 AM
11
11
52
AM
PDT
we *know* that symbols and syntax are being utilized, and we know what they specify. Then you can answer my question about how to distinguish a sequence that is one base pair from being functional, from a collection of randomly generated sequences. After all, that's one of the characteristics of syntax, recognizable words and sequences of words.
Petrushka
December 15, 2011
December
12
Dec
15
15
2011
11:48 AM
11
11
48
AM
PDT
Dr Rec: By now, it should be clear that you are imposing the a prioris, not me. Let's look at your:
I) Use of Durston’s, or any related metric imposes a post-hoc specification-a design in the search for design. Look at the tables of Fits. There is an estimate based of the length, and number of sequences. But sequences of what? A post-hoc specified design
Really, now. A protein family is observed, and its variability while retaining function is used to quantify the info in the AA sequence. We have a macro-observable state that asks only: does it do job X in living systems. That is more than sufficiently independent. The redundancy in the strings reduces the bit value from 4.32 per AA residue. After the reduction, the number of functional bits is totted up. A comparison to threshold then tells us what you obviously do not wish to hear: a functional family that isolated in AA string config space that does a job that specific to the string sequence, is not likely to have been come upon by blind processes. So we see a selectively hyperskeptical objectopn. Only problem, this is the same problem as explaining functional text on blind forces. Cicero could spot the problem c 50 BC, and so can we today, providing we are not blinkered by materialist a prioris. GEM of TKIkairosfocus
December 15, 2011
December
12
Dec
15
15
2011
11:47 AM
11
11
47
AM
PDT
Gee, when you put it that way drawing a trillion atoms from the deck and getting a sea urchin doesn't sound so strange. I can't say that DNA code symbolically representing proteins and regulations that produce an egg and grow it all the way into a reproducing sea urchin is specialized because 'I have no theory behind the distribution of function?'
You cannot independently determine a sequence that is one base pair from functionality from a purely random sequence. ... If you cannot distinguish a degraded functional sequence from a random sequence, you cannot assign bits of information to it.
Wow, what a corner you are painting yourself into. You're saying that if I can randomly substitute a gene in sea urchin, not knowing which gene, and still have a sea urchin, that means that I have know way of knowing whether or if there was ever any functional information to start with. Except, you know, the sea urchin. You seem to be reasoning that when determining whether a sequence is specified or not, we should disregard the output of that sequence if we don't precisely understand the correlation of the sequence to the output. Because.. why? The visible, undeniable evidence that the sequence does correlate to a functional output is inconvenient and your case makes more sense without it? Let's put this another way. If you give instructions to a man and he translates them in language you don't understand to a second man who then follows most but gets a few things wrong, you cannot ascribe any functional information to those translated instructions because you don't know which part was degraded. Am I misunderstanding? To repeat, you said,
If you cannot distinguish a degraded functional sequence from a random sequence, you cannot assign bits of information to it.
How does your statement not apply in the above example? You're asserting something that's ridiculous if you just think about it for a minute.ScottAndrews2
December 15, 2011
December
12
Dec
15
15
2011
11:42 AM
11
11
42
AM
PDT
An unspecified arrangement of DNA molecules will result in absolutely nothing, every time, no exceptions. That a given arrangement results in the components of cells, the cells themselves, all of their functions, and their assembly into greater units which in turn perform additional functions is evidence that the arrangement is specified. You can dispute this and reason that I’m still arbitrarily applying the specification post-hoc. But then it becomes your position that there really is no difference between a sea urchin and a pile of lifeless, functionless proteins. It’s arbitrary, splitting hairs.
Again, you are assuming that there is some fundamental difference between life and other arrangements of matter. Just because human beings belong to the category of the former, does not mean that anyone or anything intended it. And so it is not warranted to say that the complex functions that result in what we call life are 'specified'.lastyearon
December 15, 2011
December
12
Dec
15
15
2011
11:31 AM
11
11
31
AM
PDT
"We also know that it is algorithmically compressed, hierarchically nested, uses multilayered encryption," We do? Could you give an example of each in biology?DrREC
December 15, 2011
December
12
Dec
15
15
2011
11:22 AM
11
11
22
AM
PDT
Point being, with said manuscript, there is at least an appearance of specification (use of syntax, symbols), whether we know what it specifies or not. With biology, we *know* that symbols and syntax are being utilized, and we know what they specify. And we also know that it is algorithmically compressed, hierarchically nested, uses multilayered encryption, and regulates machinery, just like any modern computer would.M. Holcumbrink
December 15, 2011
December
12
Dec
15
15
2011
11:12 AM
11
11
12
AM
PDT
Is the Voynich Manuscript specified? Does it contain information, or is it just gibberish?
We don't know, but we do know that SOMEONE wrote it, with the intent to do so, gibberish or not.M. Holcumbrink
December 15, 2011
December
12
Dec
15
15
2011
11:04 AM
11
11
04
AM
PDT
i mean a bases' alphabet (of the type 'G' 'GGA' 'TT' etc..)krtgdl
December 15, 2011
December
12
Dec
15
15
2011
11:02 AM
11
11
02
AM
PDT
That sounds like Bible code to me. You can write a function to translate anything into anything else.Petrushka
December 15, 2011
December
12
Dec
15
15
2011
10:43 AM
10
10
43
AM
PDT
and if a 'genetic alphabet' existed allowing to translate the genome into an unknown poem, would you deem DNA as designed?krtgdl
December 15, 2011
December
12
Dec
15
15
2011
10:36 AM
10
10
36
AM
PDT
"Using Durston’s Fits values — functionally specific bits — from his Table 1, to quantify I, so also accepting functionality on specific sequences as showing specificity giving S = 1" I) Use of Durston's, or any related metric imposes a post-hoc specification-a design in the search for design. Look at the tables of Fits. There is an estimate based of the length, and number of sequences. But sequences of what? A post-hoc specified design. Lets do an easy one-Insulin. They list 419 sequences. I get over 900 in a blast search. The difference is where the boundary is drawn-do we include "insulin-like" proteins that perform the same function? Insulin is a human concept with a homology cutoff based on similarity to human insulin. Insulin-like proteins that perform related functions get lopped out. So what has happened here: a) From a wide range of sequences, not just a function, but a FORM (a design) has been specified. Post-hoc. b) A cutoff based on similarity has been made c) From this narrowed range of sequence space, a metric is made that is then used in turn to detect design. This ignores: I) Evolution has no target, where this metric narrowly specifies one II) Other insulin and insulin like sequences that perform the same function have been lopped out III) This isn't a true mapping of functional space-other proteins not yet sequences, or that don't even exist in nature that could substitute Take RecA, as another example. They only consider homologues above a certain detection threshold. They only consider RecA, but not other recombinases. This ignores other proteins that can substitute in functional space. Unlike insulin, taking the whole of RecA also ignores it has functional subparts, which would have lower fits. So there is a specification in the search for design. You're querying design with design. II) Sequence-based specificity metrics are fundamentally flawed A) The method estimates the functional portion of sequence space by known sequences that code for a given protein. There is going to be a significant underestimate of functional space in the connection of it to sequence space i) Evolution has not explored all of sequence space ii) Humans haven't sequenced all of biology iii) Most sequences are from evolutionarily related organisms iv) Most sequences are of unknown function, and therefore there could be many islands of sequence space which perform the same function B) Conversely, in the case of a small, novel family, the sequence space is very low, and the method would detect design. Basically anything new (sequence space=1) is declared design. C) It has no consideration of evolutionary paths. The number of fits for some whole domains and enzymes is well below the universal probability bound. If two of these recombine (which is almost a given in some organisms) the number would be pushed over the universal probability bound. So apparently natural+natural undergoing a natural (and highly probable) process yields design detection. III) As the previous example shows, what really matters isn't the state, but the change of fsc over time. The paper states repeatedly "The measured value ? of a biosequence S can change over time with mutation events." Figure 1 is "Changing measure of FSC over time." So the relevance to ID is finding an event where fsc increased over the universal probability bound at once. Which you simply don't have. Design has never been detected in biology.DrREC
December 15, 2011
December
12
Dec
15
15
2011
10:29 AM
10
10
29
AM
PDT
1 2 3 4 5

Leave a Reply