|
Author
|
Topic: Design Inference Game II
|
FR_GR
Member
Member # 1085
|
posted 21. January 2004 01:31
Very interesting discussion going on here. I'm almost sorry I chose the pictures that I did, as I'd do it all quite a bit differently today than I did last year. Ah, well....
IP: Logged
|
|
Cre8ionist
Member
Member # 140
|
posted 21. January 2004 08:23
Fortunately Pim, Dembski has given us some guidance here. Speaking of Dawkins he writes:
quote:
METHINKS IT IS LIKE A WEASEL
(he considers only capital Roman letters and spaces, thus 27 possibilities at each location in a symbol string 28 characters in length).
If we tried to attain this target sequence by pure chance (for example, by randomly shaking out Scrabble pieces), the probability of getting it on the first try would be around 1 in 10^40, and correspondingly it would take on average about 10^40 tries to stand a better than even chance of getting it. Thus, if we depended on pure chance to attain this target sequence, we would in all likelihood be unsuccessful. As a problem for pure chance, attaining Dawkins' target sequnce is an exercise in generating specified complexity, and it becomes clear that pure chance simply is not up to the task. (Technically, Dawkins' target sequence is not long enough for its probability to fall below the 1 in 10^150 universal probability bound or correspondingly for its complexity to surpass the 500-bit universal complexity bound. Dawkins' target sequence therefore does not qualify as complex specified information in a strict sense--see sections 2.8 and 3.9. Nonetheless, for practical purposes the complexity is sufficient to illustrate specified complexity.) pg 181 NFL Bold mine, formatted for forum
Now, I must rely on Dembski, as he is an expert in the field. If a string of 28 characters is enough for him, one of 45 characters is enough for me, in an informal sense.
However, this is not the point again. Surely you understand that I could simply put a picture up with enough text to force the complexity to surpass the UPB. But do we even need to go there? For purposes of our informal discussion here 45 characters will do, and the Earth is round (even though as Frances eagerly points out, it's really an "oblate sphere" ![[Wink]](wink.gif) ).
Anyway, let me make one more point before we move on.
Dembski's criteria for detecting designsignatures isn't being used any differently than the Magnetite Assay for Biogenicity (MAB) (the criteria for detecting biosignatures) would be used. Just because you know the origin of the magnetite, and you know it's from a biological source doesn't invalidate the test. In order for the test to be valid, it must work in cases where you know the origin as well as those you don't. If it isn't useful where you know the origin, how could you be confident in cases where you don't?
Likewise, if Dembski's criteria isn't useful on specified complexity that we know the origin of, we couldn't have confidence that it worked in cases where didn't know the origin. Conversely, if it does work in all cases where the origin is known, we can then use the criteria on cases where the origin is unknown with full confidence. Until there are "real" exceptions, and by real, I mean exceptions which silence Dembski (so far the exceptions have been debatable), I personally have full confidence in the criteria. Learning to use the criteria is what's critical, and I applaud this thread for its contribution to that effort. I certainly don't have its use mastered, and am grateful to this forum for a chance to voice my opinion.
I hope my point about picture selection hasn't been lost here, for a picture could easily be worth a thousand words.......Cre8
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 21. January 2004 10:23
Cre8ionist wrote quote:
However, this is not the point again. Surely you understand that I could simply put a picture up with enough text to force the complexity to surpass the UPB. But do we even need to go there? For purposes of our informal discussion here 45 characters will do, and the Earth is round (even though as Frances eagerly points out, it's really an "oblate sphere" ).
Sure, but the 'game' as I understood it from the early discussion is to apply the design detection apparatus formally to the various stimuli , not informally. Quoting again from general_re via FRDiamond, quote:
And so in order to ensure that the filter itself is the thing being tested - and not simply one's knowledge of the sorts of things that humans make - requiring the application of the filter seems reasonable enough.
That must include the UPB.
Second, about "specification." Precisely what is the specification for the letter sequences of #1? Remember, a specification is an independently given pattern, a "detachable" pattern. Cre8 has not identified that detachable pattern of the letter sequence of #1 beyond saying something about it being in English, which depends on our background knowledge of English words. But so are a whole lot of other 47-character letter sequences. Without a "specification" that singles out just the observed pattern, it's the improbability of getting any member of that class of 47-character sequences composed of legal English words that is the "complexity" of the object. In order for the (im)probability calculation to have a numerator of just 1, the "specification" - the detachable pattern - has to limit the permissible patterns to just that specific sequence, and no such specification has yet been supplied. Any sequence of 47 letters and spaces has the same probability of occurrence; what makes this one special? No specification that delimits the universe of numerators to just that sequence has been provided beyond identifying its language as English. So far, the target has been drawn around the arrow.
Is our background knowledge that the letter sequences of #1 form legal English words (not sentences) a sufficient "specification"? If so, then the corresponding "specification" of the bacterial flagellum is no more than "the chemically and physically permitted aggregations of N proteins," where N = the number of proteins comprising the flagellum.
This, of course, is where "function" enters. In order to delimit the class of patterns meeting a specification, one must invoke function.
RBH
[ 21. January 2004, 10:40: Message edited by: RBH ]
IP: Logged
|
|
|
|
Pim van Meurs
Member
Member # 541
|
posted 21. January 2004 11:58
RBH points out the difference between an actual design inference based on the requirements set forth by Dembski and how we tend to infer design in known instances. Even if the complexity does not seem to reach the required bounds, we still are tempted to make a design inference. Of course accepting a much lower bound is likely to make the design inference even more succeptible to lack in our knowledge or in other words false positives.
Perhaps Dembski could comment on Creationist's arguments?
IP: Logged
|
|
David Garrett
Member
Member # 759
|
posted 21. January 2004 12:08
RBH wrote:
quote:
Well, now, that's not always the case. Consider the Voynich manuscript. It's been around since the 17th century, and people still don't know if it in fact bears a language. While it's catalogued in the Museum of Hoaxes, there are still those who believe it will be someday be deciphered. But at this stage, we don't even know if it's a form of language in spite of the craftsmanship of the marks.
Your point is well taken although I find it interesting to point out that our assumptions about the manuscript still point to ID. Even though we can't crack the language, we still assume it to be intelligently designed even though we might chalk it up to a hoax rather than entertaining any thoughts of a non-ID explanation of it.
Before moving on I would like to bring up an issue about the "proximate function" (PF) of the basketball. Left to its own devices and the forces of nature, a basketball does nothing but succumb to the forces of gravity and sits doing nothing. However, if we were an alien race given a homework assignment of studying basketballs on Earth, we would find that there is high probability that we find basketballs at basketball games. We would also find that they tend to be bouncing at these games. What does this say about inferring design? Especially if we see the primates (whether we know the level of intelligence of the primates or not) playing with them. We most likely would use the basketball and the game to infer the primates intelligence. I only bring this up to point out how important the surrounding environment is on PF and inferring design.
FRDiamond wrote:
quote:
I can provide information about the chemical properties of the pictures if that will help.
I would like to move discussion towards the second item. Could you please provide the material composition of the second item?
Dave
IP: Logged
|
|
|
|
|
|
|
|
Cre8ionist
Member
Member # 140
|
posted 21. January 2004 15:27
RBH writes:
quote:
Without a "specification" that singles out just the observed pattern, it's the improbability of getting any member of that class of 47-character sequences composed of legal English words that is the "complexity" of the object. In order for the (im)probability calculation to have a numerator of just 1, the "specification" - the detachable pattern - has to limit the permissible patterns to just that specific sequence, and no such specification has yet been supplied. Any sequence of 47 letters and spaces has the same probability of occurrence; what makes this one special?
Good question, let me fall back on Dembski's example from "Contact." Suppose the following set of zeros and ones is the next picture we display. If you've followed Dembski's work you know it's not just random. It's the sequence of 1,126 beats and pauses the SETI researchers received from outer space in the movie. The sequence represents the prime numbers from 2 to 101. If you're familiar with the film you also know they took this as "decisive confirmation of an extraterrestrial intelligence." Were they guilty of drawing the target around the arrow? How many possible combinations here? 2^1126! Out of those how many structured possibilities? (Recall that the researcher who discovered the sequence said "This isn't noise, this has structure.") Yet the probability is small enough here to detect design, which is the point.
My challenge to you is, run the following picture through the filter? But to make it interesting let's say the Mars rover Spirit found this not as 1s and 0s, but as dots and dashes etched somewhere on the Martian surface and sent the picture back for us to examine.........What's your verdict, designed or natural?...Why?........Cre8
110111011111011111110111111111110111111111111101111111111111111101111111
111111111111011111111111111111111111011111111111111111111111111111011111
111111111111111111111111110111111111111111111111111111111111111101111111
111111111111111111111111111111111101111111111111111111111111111111111111
111111011111111111111111111111111111111111111111111111011111111111111111
111111111111111111111111111111111111011111111111111111111111111111111111
111111111111111111111111110111111111111111111111111111111111111111111111
111111111111111111111101111111111111111111111111111111111111111111111111
111111111111111111111101111111111111111111111111111111111111111111111111
111111111111111111111111011111111111111111111111111111111111111111111111
111111111111111111111111111111110111111111111111111111111111111111111111
111111111111111111111111111111111111111111110111111111111111111111111111
111111111111111111111111111111111111111111111111111111111111110111111111
111111111111111111111111111111111111111111111111111111111111111111111111
111111111111111101111111111111111111111111111111111111111111111111111111
1111111111111111111111111111111111111111111111
[ 21. January 2004, 18:28: Message edited by: Cre8ionist ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 21. January 2004 16:21
Given the last picture, I would like to add the following one as well.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 22. January 2004 03:20
Those two are essentially arrays of hexagons. Although I am missing details of what each one is composed of exactly (or I am pretending that I am), hexagons are easily created by a regularity. You need a process that creates a unit cell of a consistent size (e.g. raindrops can be fairly consistent) and packs them together in a planar array (e.g. the raindrops fall through a film of protein into oil and sink to the bottom of a jar). Because there is a plausible regular mechanism to explain hexagons in general, it seems that there would be too little information in a picture to confidently derive a probability bound of less than 10^-150. So one would not conclude design with either of those pictures, using Dembski's method.
One could make a similar argument for the snowflake-umbrella, now that I think of it, so you could not safely conclude design there either.
The basketball is more confusing, as probabilities are harder to estimate with confidence. Here we cannot safely say whether or not one cannot safely conclude design.
IP: Logged
|
|
gregorythegrey
Member
Member # 1088
|
posted 22. January 2004 04:25
Greetings ISCID Board,
First time back again. I’m a little late for this thread topic, but since still less than 5 pictures have been posted...
Nice game, though...however, I’m curious about one thing in particular with regard to applying Dembski’s theoretical EF in order to make design inferences. When does one distinguish between human-made things and non-human-made things?
quote:
"...that question of who the designer might be, because our intuitive judgments of what is designed are inevitably colored by our background knowledge of the only kind of designer we know - human beings." - Evan (15-01-04)
For example, in the video Unlocking the Mystery of Life, examples are given to support IDT or the IDM that include Archaeology, SETI, the familiar mousetrap, animal-shaped bushes (like in Edward Scissorhands), the statues of Easter Island and Mount Rushmore. Yet these things are all human-made and very different from an organic or biological example. The jump from basketball to flagellum, of course, is not so easy to maintain and would seem to require very different observers, if not very different rules and methods. Mathematics and probability do not seem as relevant to human-made things. What do others think?
To me (as a recently retired university b-ball player), it is obvious that the basketball was ‘designed’ since James Naismith invented the game in 1891. (ie. Basketballs do not commonly descend from the original!?) This link actually gives the U.S. patent number from 1929. Basketball was invented and the spread in popularity of the game led to a diffusion of new basketball ‘designs’ within the specifications laid out originally.
Side point of potential interest: there is also a basketball named “Evolution” (rather than 'Solution,' which may turn out to somehow be related) which was made about five years ago by Wilson Jet. Quite a coincidence! Nonetheless, I would argue with anyone, any day, any time that basketballs, my friends, definitely DO NOT EVOLVE into being, or into new models, shapes (well, round should still do) or forms by themselves. Even if you give them new names. That is, humans are employed by sporting equipment companies who ‘invent’ new ball ‘designs’ and brand them.
In any case, it seems others are confused by the human-made, non-human-made distinction as well and that is why I’m asking this question.
With kind regards,
Gregory the Grey
Appreciated words:
quote:
“And so in order to ensure that the filter itself is the thing being tested - and not simply one's knowledge of the sorts of things that humans make - requiring the application of the filter seems reasonable enough.” – FRDiamond (16-01-04)
quote:
“In truth, though, the points about historical knowledge and the qualities of the designer turn out to be moot, for the most part - the bulk of the objects selected for this initial round were not what we might call ‘obviously’ or ‘trivially’ designed by human agents, and so the players had little choice but to operate via the filter, as background knowledge of the ‘frying pans are made by humans’ sort wasn't really available. At least, that was the idea.” – FRDiamond (16-01-04)
quote:
“The basketball is more confusing, as probabilities are harder to estimate with confidence. Here we cannot safely say whether or not one cannot safely conclude design.” – Rex Kerr (22-01-04)
[ 22. January 2004, 07:38: Message edited by: gregorythegrey ]
IP: Logged
|
|
Cre8ionist
Member
Member # 140
|
posted 22. January 2004 07:52
Hi gregorythegrey,
I'm trying to figure out how a SETI example is necessarily a human example? Anyway, what all the objects/events need to share that the EF can identifiy is CSI or, more commonly SC. If an object/event doesn't display the characteristics of SC, and only these characteristics, it will not return a positive result from the filter. However, as I pointed out earlier, this doesn't mean the object/event isn't designed, it only means that the filter is useless in the determination.
Also, picture selection here appears to be a bit one-sided. Obviously, someone from the design side could pick easier examples, as I showed above with the as yet unaddressed "Contact" example. In this game picture selection is everything. I personally think that you would start out with easy examples and move to more difficult ones, as establishing the filter's usefulness in some early examples would seem to be of some importance.................Cre8
[ 22. January 2004, 08:08: Message edited by: Cre8ionist ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 22. January 2004 23:43
Rex: Those two are essentially arrays of hexagons. Although I am missing details of what each one is composed of exactly (or I am pretending that I am), hexagons are easily created by a regularity.
But that is because of your background knowledge about nature and hexagons. But lets assume for the purpose of the argument that we only had observed so far designed hexagons for instance by bees? And then suddenly we observe the hexagons in my example. Assuming that we can eliminate chance and regularity we now have to decide if the observed hexagons exhibit specification. Their intelligently designed counterparts seem to suggest that there is a specification, is there complexity? Well here things get complicated because as usual, the calculation of complexity is far from trivial except for the most trivial examples. But what is the probability of hexagons, many of them being repeated given my understanding of nature? I would argue that this could be a good candidate for a false positive.
IP: Logged
|
|
|
posted 21. January 2004 11:01
Printer-friendly view of this topic