Guest Post: Design Detection

_{Barry Arrington

December 10, 2014

Intelligent Design}

Share: Facebook; Twitter; LinkedIn; Flipboard; Print; Email

Paul Giem provides the following guest post today:

The following three pictures were made to represent trays with 560 coins with either white (heads) or black (tails) showing. At least one of them was created by shaking coins and then spreading them out on a table (actually multiple shakes of 20 or so coins) and copying the pattern of heads and tails produced. Which one or ones are they, and why? Were the ones, if any, that were not done by this process designed, and if so by whom, and using what method?

Comments

Paul Giem, Wonderful post and great thread. Very educational. Thankyou.logically_speaking_{December 13, 2014
December
12
Dec
13
13
2014
06:25 AM
6
06
25
AM
PDT}

Zac said, if it depends on the metaphysics of the observer, it is neither objective nor scientific. I say, Metaphysics is by definition above and beyond physics. There are no objective facts that are above and beyond Metaphysics. Worldview trumps everything. As far as being not scientific. Are you claiming that you must have a materialistic worldview to do science? Or are you claiming that the only evidences that can be considered scientific are those that don't challenge your materialistic worldview? peacefifthmonarchyman_{December 13, 2014
December
12
Dec
13
13
2014
06:22 AM
6
06
22
AM
PDT}

E.Seigner: I have been saying nothing else but that context is everything, and context includes the metaphysical presuppositions of the observer. According to Dembski in "Specification: The Pattern That. Signifies Intelligence", there is a methodological determination. Indeed, if it depends on the metaphysics of the observer, it is neither objective nor scientific.Zachriel_{December 13, 2014
December
12
Dec
13
13
2014
04:01 AM
4
04
01
AM
PDT}

Mung #19
Mark, we get people in here all the time who assert that design is not detectable. E.Seigner comes to mind as a recent example.
You are referring to this statement by Mark: "No one is denying the ability to detect if something is designed." But note what immediately followed: "The dispute is how and when can you detect design, and in particular do you have to know something about the context in which it was created to detect if it was designed or can you tell simply from some intrinsic property of the object itself..." which is precisely what the debate is about. I have been saying nothing else but that context is everything, and context includes the metaphysical presuppositions of the observer. Mung #19
Note also that all of a sudden who designed #3 doesn’t seem to be an issue. Hidden and mysterious powers of the putative designer seem to be irrelevant all of a sudden. The motive? Also not relevant. Their religion? Not relevant. Do they have a Wedge document? Not relevant. Are they a secret admirer of the Discovery Institute? Not relevant.
Note also that, partly because of lack of answers to these questions, the exercise of the OP is scientifically irrelevant and philosophically uninteresting. All answers seem to be applicable to the OP. Yes, all three are designed because they are all made of designed coins or they all have a structured distinction of black and white sides. No, none of the three is designed because they are computer-generated images consisting of bits and bytes in virtual webspace, and any perception of structure, colour, words, etc. is objectively not there in the fundamental constituents of the images, but read into them by the observer. Thus the OP is scientifically irrelevant and philosophically uninteresting, until further relevant distinctions and purposes have been clarified.E.Seigner_{December 13, 2014
December
12
Dec
13
13
2014
01:37 AM
1
01
37
AM
PDT}

Paul: Very good work! I completely agree with all your comments. :)gpuccio_{December 12, 2014
December
12
Dec
12
12
2014
10:22 PM
10
10
22
PM
PDT}

PS. image 3 is specified by language and meaning (although ironic), whereas image 2 is specified by function. So don't be surprised if you don't immediately see the meaning in a long string of DNA bases.Paul Giem_{December 12, 2014
December
12
Dec
12
12
2014
09:39 PM
9
09
39
PM
PDT}

To all, My own thoughts (at present) on the subject of design detection are as follows: We recognize design when we find a pattern that makes sense from a design perspective and is not required by law, not likely by chance, and not made likely by some combination of chance and law. Law is a possible explanation, and in the case of 500 coins all with the heads up, although almost all chance explanations can be ruled out, the law explanation of two-headed coins cannot be ruled out, and in fact explains the phenomenon perfectly. That is why my illustrations avoided that example. Pure chance cannot explain images 2 and 3, but can explain image 1, and therefore we can assume image 1 is random, but can assume with well over (1 - 2^-500) probability that both image 2 and image 3 are not due to chance. I have not seen any law-like process that causes the binary digits of pi to be laid out, and neither have I seen any law-like processes that cause English words to be laid out in binary code, that have not been designed (this requirement excludes computer programs). In the absence of such law-like processes, or law-and-chance like processes, that could explain the appearance of design, the appearance is best taken at face value. This evaluation would have to change if such processes were discovered. But science has to deal with the known, and it is reasonable to draw the apparently obvious conclusion until such time as its obviousness comes under challenge. One can claim that in the case of image 3, the target area is large. This is true. It is certainly larger than the area of the pi target (see my comments in #61). The pattern of heads and tails could have been reversed (increasing the target area by a factor of 2), the top half (or the bottom half) could have been reversed (another factor of 2), a different font could have been selected (perhaps a factor of 1,024 or 2^10), different (short) English words could have been chosen (perhaps a factor of 1,048,576 or 2^20), or a different written language could have been chosen (perhaps a factor of 16,384 or 2^14, although to be fair, English is the second most common language, and the most common on the internet), and there is one mistake which could have happened in any of 560, or roughly 2^9 spots. This makes our target sequence somewhere in the neighborhood of 2^55 sequences, thus increasing the probability of hitting the target randomly to roughly 1 in 2^505 (or less). The positive evidence for design, plus the extremely low probability of finding that target by either chance or any known chance/law combination, rationally supports our instinctive reaction that image 3 must have been designed. But perhaps someone else has a better explanation of how we (accurately) detect design.Paul Giem_{December 12, 2014
December
12
Dec
12
12
2014
09:34 PM
9
09
34
PM
PDT}

PaV (#58), I fully understand your inability to reply rapidly. I can be out of communication for days at a time, as I have a real job. Hopefully our communication is worth waiting a few days for, when necessary. I disagree on the ability to be sure that image 2 is designed. This is not just because I happen to know it was designed, as I was the designer. It is also because I know how hard it is to get the first 560 binary digits of pi (which is itself the pattern that emerges). If one is randomly guessing, the probability is 1 in 2^560, which is roughly 1 in 3.8 x 10^168. That number is well above the Dembski universal probability bound, let alone the (IMO more reasonable) Dawkins reasonable bound of 10^50. One could reduce the probability by saying that black could stand for 1 and white could stand for 0, but that only multiplies the probability by 2. One could further increase the probability by saying that we could start at the right (factor of 2), or perhaps go back and forth (factor of 2), or start at the bottom (factor of 2), or go up and down (factor of 2 for all variations). We could even spiral in or out or in (much less than a factor of 2 at this point). After trying all these variations we will have reduced the odds to somewhat worse than 1 in 2^554, or 5 x 10^166, which is trivial odds. One can claim that there are other special numbers, such as 2*pi, pi/2, and pi/4, but they all have the same binary digits; one simply moves the ~~decimal~~ binary place over. Perhaps one can simply divide by 3, multiply by 3, divide by 5, or multiply by 5, and still get a special number, but this only raises our odds by 5. Besides, the fact that we hit dead on a very special number makes these hypotheses seem ad hoc. One could argue for e, or perhaps the square root of 2 (= the square root of 1/2 in binary except for the ~~decimal~~ binary place), but this only raises the probability by a factor of 3. This is still way beyond the Dembski universal probability bound. Even if there is somehow an error in the transcription (point mutation, or single frameshift deletion, or addition of 1 or 0), this gives us about 2,237, or roughly 2^11, rise in probability. This is peanuts in comparison to 2^560, or even 2^554. That is why, after I had created image 1, I didn't even bother to check to see if the binary digits of pi, or any of those other numbers, could be found in it. I am mildly surprised to note that Silver Asiatic found a variant (not perfect--there are 2 white separators at one point) of 3.141 at the beginning. I would have been totally astounded if the pattern had continued to the next 2 digits, 5 and 9. While I agree with cklester (#54) that one cannot be 100% sure that #2 is designed, after the pattern is pointed out and verified, one can be 99.9999[160 more 9's]% sure that it is designed, and that would be good enough for me. Also, without knowing the backstory, one could not be sure that #1 was not designed (or at least the heads-tails string was not designed), it seems reasonable in the absence of convincing evidence otherwise to accept it as random.Paul Giem_{December 12, 2014
December
12
Dec
12
12
2014
08:46 PM
8
08
46
PM
PDT}

PaV, I've no idea how life started, but you seem to think there is a way to calculate CSI or something like it for the origin of life. I don't really see how -- neither the chance hypothesis nor the target-space seem well defined to me.wd400_{December 12, 2014
December
12
Dec
12
12
2014
07:01 PM
7
07
01
PM
PDT}

Me_Think (#48),
Gary S. Gaulin @ 47 No idea why you referred to Pseudorandomness. Paul Giem’s idea is to show 1 is random, 2 is complex and 3 is Complex and specified.
You are close. 1 is random, and 2 and 3 are both complex and specified. It is just that 3 looks non-random to the naked human eye, whereas 2 does not, whereas, unless I have made a mistake, 2 is actually slightly less random than 3. Incidentally, to a computer reading the rows and putting them into a linear array, once a hypothesis has been suggested, it is algorithmically easier for the computer to check and find 2 to be non-random than 3.Paul Giem_{December 12, 2014
December
12
Dec
12
12
2014
06:41 PM
6
06
41
PM
PDT}

Paul Giem:
Now, the question I have from that is, is there anyone out there that wants to make the case that number 2 was made by shaking coins and spreading them out, now that we have this kind of a match to pi? If so, or if not, why? Do you detect design? If so, or if not, why?
Paul, I have very limited time to discuss this. Over the next two days, I will likely not be able to get back to you. Here goes: I have difficulty concluding that this is design for one simple reason: in a certain sense there is NO "pattern" here. You have a string of numbers that, combined, are the 'representation' of Pi. But pi is irrational and quasi-random. While this fulfills all of Dembski's requirements, it's randomness makes the design conclusion an uneasy one. Dembski's example of 500 binary digits which, in ascii, form the first 100 prime numbers is different. There one has the "pattern" of 'prime numbers' emerging. With pi, OTOH, no such "pattern" emerges. Whereas I would be comfortable concluding design in Dembski's case, in yours I would hesitate. Obviously, #3 has a "pattern."PaV_{December 12, 2014
December
12
Dec
12
12
2014
06:08 PM
6
06
08
PM
PDT}

You are now talking about the origin of life, so to use CSI you would have to define a minimal-replicator/metabolsim as the target.
It seems to me that you want it both ways; i.e., you simply dismiss the chemistry as being an issue for OOL, which, once solved, puts us in a different position where "non-random" effects might come into play. Yet, if "random" forces need "non-random" effects to bring about evolution, then how, unaided by "non-random" effects, did life come about in the first place? You just can't have it both ways. OOL is vastly more complicated a 'trick' to accomplish than is even 'macroevolution.' Are you a believer in "panspermia" as was Sir Fred Hoyle? But even panspermia doesn't help. It's biology version of the 'multiverse,' just pushing things farther and farther back in time. IOW, you give yourself an "eternity." Well, in an infinite amount of time, infinite odds can be overcome. This renders science meaningless.PaV_{December 12, 2014
December
12
Dec
12
12
2014
05:57 PM
5
05
57
PM
PDT}

Silver Asiatic (#55), The reason why you are not finding a binary representation of pi in image #1 is because there isn't one. That is the one created by shaking coins (multiple times) and lining them up and recording the heads and tails. It would take tremendous luck to get pi to be represented in a random string of ones and zeros. It is interesting that you were able to find 3 black, then 1 black, then 4 black, then 1 black at the beginning. That is unusual, especially when another image has an exact (or nearly exact--someone should check my work) binary representation of pi. However, it is not quite as unusual as one might think. In a linear array, half of all black series should be single (giving 1 by your method), and another quarter should be 2's, an eighth should be in a series of 3, and a sixteenth should be in a series of 4. The series stalls out at 3.14159 because a 5 is fairly improbable (1/32), and 9 is extremely improbable (1/512) and not likely to happen even once in our picture. By my calculations the probability of getting 3141 as the first 4 digits by your method is (1/2*1/8*1/2*1/16=) 1/512, a small but not astronomically small probability. The fact that one cannot continue the series reduces the probability that this was in fact designed. As the person who created that series, I can assure you that in fact image #1 does not have the series of black and white circles designed. It is rather reproducing a series that uses a method that reliably gives random results. If you want to find a binary representation of pi, check image #2.Paul Giem_{December 12, 2014
December
12
Dec
12
12
2014
05:53 PM
5
05
53
PM
PDT}

In image #1 I see this much of the sequence in black 3.141 left to right just counting. I'm not finding the binary representation though but I haven't checked every direction.Silver Asiatic_{December 12, 2014
December
12
Dec
12
12
2014
01:49 PM
1
01
49
PM
PDT}

I, as an English speaker, can feel confident in saying "3 is designed," but I cannot determine if 1 or 2 is designed (could be a code in binary or some other pattern in a code with which I'm not familiar). So, the least I can ever say is, "It doesn't look designed." To say something is "not designed" will always be an argument from ignorance, won't it? Just like "junk DNA" or "vestigial organs." It will always be possible that we just don't have enough info, yet. Can I ever, with 100% confidence, say, "It is designed?"cklester_{December 12, 2014
December
12
Dec
12
12
2014
01:19 PM
1
01
19
PM
PDT}

kairosfocus (#49),
I clip the first several blocs of binary pi: 00100100 00111111 01101010 In block 2, I see 4 B, 6 W in a bloc, then BWWBWB, which fits the pi pattern.
Actually, if you look at the first line of block 2, you not only have the bolded part of the sequence matching, but also the next two digits, and the first 6 digits, which leaves only the first 2 digits of pi (11), which match the first 2 coins assuming white is 1 and black is 0. You can then see how far out it goes. The sequence either exactly, or very closely, matches the ones and zeros of binary pi. Now, the question I have from that is, is there anyone out there that wants to make the case that number 2 was made by shaking coins and spreading them out, now that we have this kind of a match to pi? If so, or if not, why? Do you detect design? If so, or if not, why? What about block 1?Paul Giem_{December 12, 2014
December
12
Dec
12
12
2014
09:19 AM
9
09
19
AM
PDT}

The first two look like a game of go between two beginners.Seqenenre_{December 12, 2014
December
12
Dec
12
12
2014
07:48 AM
7
07
48
AM
PDT}

Me_Think:
Gary S. Gaulin @ 47 No idea why you referred to Pseudorandomness. Paul Giem’s idea is to show 1 is random, 2 is complex and 3 is Complex and specified.
The following formed a "pseudorandom" string:
Go to the website, and get the first 20 digits of the binary form of pi: 11.00100100 00111111 01 Transform them into heads and tails: HHTTHTTHTTTTHHHHHHTH or conversely TTHHTHHTHHHHTTTTTTHT Do either of these match any sequences in number 1 or number 2? If you find a match, see how far out it goes.
Gary S. Gaulin_{December 12, 2014
December
12
Dec
12
12
2014
05:08 AM
5
05
08
AM
PDT}

PS: One of the many fallacies that surround the debates over design is the notion that the design inference process ought somehow to be a universal decoder and puzzle solver. But, we already know that algorithms are very limited, so we have no right to expect or demand a universal decoder or puzzle solver. In answer, I say, first we identify or recognise a pattern or function depending on wiring diagram interaction and specific configuration, and then we address the complexity-specificity criterion, before we may reasonably infer design on signs such as FSCO/I. Of course, it turns out that in a great many relevant contexts, that is not hard to do, indeed we actually saw recognition of FSCO/I in the 1970's before the significance of that was drawn out and a technical design inference on FSCO/I was made.kairosfocus_{December 12, 2014
December
12
Dec
12
12
2014
04:12 AM
4
04
12
AM
PDT}

PG: I clip the first several blocs of binary pi: 00100100 00111111 01101010 In block 2, I see 4 B, 6 W in a bloc, then BWWBWB, which fits the pi pattern. Of course, as noted, the distribution of 1's and 0's in pi will reflect the lack of correlation between the place value notation scheme, and the ratio of circumference to diameter of a circle. So, the distribution of digits should mimic a random one. But, when one discerns a separate functionality or pattern marking a specification, that then points to complexity plus specificity thus design. KFkairosfocus_{December 12, 2014
December
12
Dec
12
12
2014
04:01 AM
4
04
01
AM
PDT}

Gary S. Gaulin @ 47 No idea why you referred to Pseudorandomness. Paul Giem's idea is to show 1 is random, 2 is complex and 3 is Complex and specified.Me_Think_{December 12, 2014
December
12
Dec
12
12
2014
03:40 AM
3
03
40
AM
PDT}

http://en.wikipedia.org/wiki/PseudorandomnessGary S. Gaulin_{December 12, 2014
December
12
Dec
12
12
2014
01:50 AM
1
01
50
AM
PDT}

anthropic (#42), Silver Asiatic (#44), and fifthmonarchyman (#45), also kairosfocus (#36), Go to the website, and get the first 20 digits of the binary form of pi: 11.00100100 00111111 01 Transform them into heads and tails: HHTTHTTHTTTTHHHHHHTH or conversely TTHHTHHTHHHHTTTTTTHT Do either of these match any sequences in number 1 or number 2? If you find a match, see how far out it goes. Box (#34), I agree with Kairosfocus (#37). That is why the impression of design, once seen, is so hard to shake. That is why I am allowing people to discover which of the other two pictures are designed, rather than just saying. I hope you all get it soon. Once you do, I think there are lessons to be learned.Paul Giem_{December 11, 2014
December
12
Dec
11
11
2014
10:28 PM
10
10
28
PM
PDT}

I would argue that number 2 also corresponds to an irrational constant. We don't know what that constant is but an algorithm that computes it could be expressed as...... What happened as Paul Giem was preparing the second sequence of coins. If we knew the constant we could reproduce the 2nd sequence exactly. There really is no such thing as random. Only apparent randomness. IMHO peacefifthmonarchyman_{December 11, 2014
December
12
Dec
11
11
2014
01:11 PM
1
01
11
PM
PDT}

anthropic 42
Thought I had it: Number one is also designed, with the value of pi: (dark coins) 3 space 1 space 4 space 1. But then the pattern didn’t fit, probably because I’m lousy at puzzles!
If you count the black dots sequentially, disregarding the spaces, it maps to pi. It would be more specified if the white spaces segmented the numbers but it wouldn't look as random.Silver Asiatic_{December 11, 2014
December
12
Dec
11
11
2014
12:55 PM
12
12
55
PM
PDT}

You are now talking about the origin of life, so to use CSI you would have to define a minimal-replicator/metabolsim as the target. Even then, I don't think anyone thinks life started as the result of nucleotides bumping into each other in a "completely random fashion" (though, again, it's not entirely clear what you mean by this phrase).wd400_{December 11, 2014
December
12
Dec
11
11
2014
12:44 PM
12
12
44
PM
PDT}

Thought I had it: Number one is also designed, with the value of pi: (dark coins) 3 space 1 space 4 space 1. But then the pattern didn't fit, probably because I'm lousy at puzzles!anthropic_{December 11, 2014
December
12
Dec
11
11
2014
12:43 PM
12
12
43
PM
PDT}

wd400: The question before us is whether or not life is "designed" by an intelligent designer, or is the result of, relatively speaking, random processes. To use the results of biology to press the case of 'non-randomness' is to argue circularly. That is why I said 'chemical' evidence. All of this is rather obvious. It is only because people don't want to accept what they fear make come from such common sense understandings that they protest against the obvious. But, perhaps, they "protesteth too much."PaV_{December 11, 2014
December
12
Dec
11
11
2014
12:20 PM
12
12
20
PM
PDT}

PaV, I said molecular biology, which, as the first word hints at, is pretty much chemistry. Even if it wasn't, why would you exclude biological evidence. Has ID regressed to trying to restore vitalism? GP, I don't think that's what PaV means, and such an understanding doesn't get us much closer to calculating the probability that this or that enzyme organ or regulatory network evolved anyway.wd400_{December 11, 2014
December
12
Dec
11
11
2014
07:47 AM
7
07
47
AM
PDT}

Box @ 34
Intelligent agent’s have a bird’s-eye view
Since you don't know about the designer how do you know he/she/it has a bird's eye view ? :-)Me_Think_{December 11, 2014
December
12
Dec
11
11
2014
05:38 AM
5
05
38
AM
PDT}

Prev 1 2 3 4 5 Next

You must be logged in to post a comment.

Leave a Reply