Fixing a Confusion

_{Jonathan Bartlett

November 30, 2016

Darwinism, ID Foundations, Intelligent Design, specified complexity}

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

I have often noticed something of a confusion on one of the major points of the Intelligent Design movement – whether or not the design inference is primarily based on the failure of Darwinism and/or mechanism.

This is expressed in a recent thread by a commenter saying, “The arguments for this view [Intelligent Design] are largely based on the improbability of other mechanisms (e.g. evolution) producing the world we observe.” I’m not going to name the commenter because this is a common confusion that a lot of people have.

The reason for this is largely historical. It used to be that the arguments for design were very plain. Biology proceeded according to a holistic plan both in the organism and the environment. This plan indicated a clear teleology – that the organism did things that were *for* something. These organisms exhibited a unity of being. This is evidence of design. It has no reference to probabilities or improbabilities of any mechanism. It is just evidence on its own.

Then, in the 19th century, Darwin suggested that there was another possibility for the reason for this cohesion – natural selection. Unity of plan and teleological design, according to Darwin, could also happen due to selection.

Thus, the original argument is:

X, Y, and Z indicate design

Darwin’s argument is:

X, Y, and Z could also indicate natural selection

So, therefore, we simply show that Darwin is wrong in this assertion. If Darwin is wrong, then the original evidence for design (which was not based on any probability) goes back to being evidence for design. The only reason for probabilities in the modern design argument is because Darwinites have said, “you can get that without design”, so we modeled NotDesign as well, to show that it can’t be done that way.

So, the *only* reason we are talking about probabilities is to answer an objection. The original evidence *remains* the primary evidence that it was based on. Answering the objection simply removes the objection.

As a case in point, CSI is based on the fact that designed things have a holistic unity. Thus, they follow a specification that is simpler than their overall arrangement. CSI is the quest to quantify this point. It does involve a chance rejection region as well, but the main point is that the design must operate on principles simpler than their realization (which provides the reduced Kolmogorov complexity for the specificational complexity).

Comments

Gpuccio: The reason why we all believe that other people are conscious is an inference from analogy.
I find this an extremely interesting insight. Really. Thank you very much. - - - -
Silver Asiatic: It also strikes me that Hume’s argument is not a reasonable approximation of what the design argument actually is. He created a straw man.
Dembski writes about Hume's criticism of the design argument: "It is this criticism that for many philosophers of religion remains decisive against design."Origenes_{December 13, 2016
December
12
Dec
13
13
2016
02:31 PM
2
02
31
PM
PDT}

Origenes
Here I would like to propose another additional premise: Q explains the presence of A, B, C and D. IOWs the presence of Q makes logical sense, given the presence of A, B, C and D.
Good addition. It also strikes me that Hume's argument is not a reasonable approximation of what the design argument actually is. He created a straw man.Silver Asiatic_{December 13, 2016
December
12
Dec
13
13
2016
01:22 PM
1
01
22
PM
PDT}

Origenes: The reason why we all believe that other people are conscious is an inference from analogy. Nobody really questions that inference. Some arguments from analogy may be feeble, but many are really strong. As for me, if I find a watch in a field, I have absolutely no doubt that it is designed. Long live Paley's argument.gpuccio_{December 13, 2016
December
12
Dec
13
13
2016
09:01 AM
9
09
01
AM
PDT}

UB: "I believe we understand and appreciate each other’s positions, and realize that we end up at the same conclusion in the end." Absolutely! And slight differences in the approach can only enrich the discussion. :)gpuccio_{December 13, 2016
December
12
Dec
13
13
2016
08:59 AM
8
08
59
AM
PDT}

Bartlett: It used to be that the arguments for design were very plain. Biology proceeded according to a holistic plan both in the organism and the environment. This plan indicated a clear teleology – that the organism did things that were *for* something. These organisms exhibited a unity of being. This is evidence of design. It has no reference to probabilities or improbabilities of any mechanism. It is just evidence on its own.
Here, it seems appropriate to discuss Paley’s famous watchmaker analogy. I am reading Dembski’s view on it, and I believe that I have a small addition to make, which strengthens Paley’s argument. The following quotes are from “No Free Lunch”, chapter 1.8, by Dembski.
According to Paley, if we find a watch in a field, the watch's adaptation of parts to telling time ensures that it is the product of an intelligence. So too, according to Paley, the marvelous adaptations of means to ends in organisms ensure that organisms are the product of an intelligence.
Hume criticized the design argument as a “feeble argument from analogy”.
Schematically, an argument from analogy takes the following form: we are given two objects, U and V, which share certain properties, call them A, B, C, and D. U and V are therefore similar with respect to A, B, C, and D. Now, suppose we know that U has some property Q, and suppose further that we want to determine whether V also has property Q. An argument from analogy then warrants that V has property Q because U and V share properties A, B, C, and D, and U has property Q. In terms of premises and conclusion, the argument from analogy therefore looks as follows: U has property Q. U and V share properties A, B, C, and D. Therefore V also has property Q. In the case of Paley's watchmaker argument, U is a watch, V is an organism, and the property Q is that something is intelligently designed. For the watch there is no question that it actually is intelligently designed. For the organism, on the other hand, this is not so immediately clear.
Dembski goes on explaining that the “difficulty with arguments from analogy is that they are always also arguments from disanalogy” — arguments from analogy can lead us astray. Next he proposes a strengthened form of Paley’s argument from analogy:
… this strengthened form of the argument therefore has an additional premise and can be formulated as follows: U has property Q. U and V share properties A, B, C, and D. There is no known instance where A, B, C, and D occur without Q. Therefore, V has property Q.
Here I would like to propose another additional premise: Q explains the presence of A, B, C and D. IOWs the presence of Q makes logical sense, given the presence of A, B, C and D.Origenes_{December 13, 2016
December
12
Dec
13
13
2016
04:35 AM
4
04
35
AM
PDT}

By the way GP, I recently passed the eight year mark since I began my project, and it has now been six years since you advised me and set me off in the right direction. I thank you for that. And now we start a new year ...Upright BiPed_{December 12, 2016
December
12
Dec
12
12
2016
09:10 PM
9
09
10
PM
PDT}

GP, Thank you for taking the time and effort to write your last two posts. We are coming at the issue from different perspectives and that distinction shows through. However, I believe we understand and appreciate each other's positions, and realize that we end up at the same conclusion in the end. I trust all is well with you, and hope you and your family have a happy and safe Christmas holiday.Upright BiPed_{December 12, 2016
December
12
Dec
12
12
2016
07:41 PM
7
07
41
PM
PDT}

UB: Some more thoughts: We can read a binary (or not binary, according to the alphabet that can be observed) sequence from any material object, provided that we can objectively observe a sequence of states that can be read linearly (it is not relevant if they are in a straight line or in a circle), and that there is a finite number of states that can be unequivocally identified and are in finite number, so that we can consider them as elements of some alphabet. Again, please note that there is nothing in that definition that assumes that the sequence of material states is designed, or that it is a representation of something. There is no assumption that semiosis is implied in what we observe. If you give a look at my example of the stone wall on a faraway planet, at post #24, you can see that we simply observe a series of signs in the stone. As far as we know, they can well be random signs left by weather events. But they are in some form of linear sequence, and they can easily be categorized as belonging to two different categories. Therefore, according to my definition, we can read a binary sequence from that object. Indeed, we can read 8 sequences, if I am not wrong, according to where we start and the direction we read. If the object is not designed, ID theory says that for none of those 500 bit sequences we will be able to define a specific function. So, if we find that one of them, any of them, a specific function whose complexity is 500 bits, we can infer that the whole objects is designed. That's why, if one the sequences we can read perfectly corresponds to the sequence of the first 125 decimal digits of pi. considering each four binary digits as a redundant code for decimal digits, then we can infer design. Now, this is IMO a very powerful example, because it emphasizes an important point: there is no restriction at all on the sequences that we ca read from material objects. The only required thing is that we can read them, and then define a function for them. In the case of pi, our function requires a symbolic code: we read the signs as binary numbers, we group them in words of 4 elements, and we read those words as decimal digits. That is certainly an added bonus, an important one, and it certainly makes the design inference much stronger, because we have not only the argument from functional complexity, but also the argument from semiosis, that are IMO the single two basic arguments that allow a design inference. In the case of the key, there is only the argument from functional complexity, and not the argument from semiosis. But the design inference is valid just the same. Let's say that the key is 500 elements long, 500 spikes and holes in some sequence. We can read that sequence as a binary digit. For all my purposes, that's more than enough to consider that sequence digital information. For example, I can well give that key to a blacksmith, and tell him to build another one, with the same elements, and respecting the same sequence. If I just showed him the single elements, and told him to make a key with 500 of them in random order, he could do that, but the key would not work. But if I give him the key, he can use it to get the right sequence. So, the key itself is a source of information, but not a symbolic source of information, because the key does not represent another object. If each single position must be right to get the function (to open the safe), then the key is the source of 500 bits of functional information. Even if there is no semiosis implied. Design can be safely inferred. Now, I will transfer the concept to an example that is much nearer to our discussions: proteins. Let's consider my old friend, the beta chain of ATP synthase. Let's say that we already have all the rest, and we need that chain to build a working molecule of ATP synthase. But we don't know the right sequence. Now, there are two different ways that we can get the right sequence. 1) We can look at the gene for that sequence, observe the sequence of nucleotides, group them in words of three, and then identify the start and end, and each single aminoacid, by translating that sequence by the genetic code, that I am assuming we know. At that point, we can synthesize the chain in our lab, because we have the correct information 2) We can analyze an existing molecule of the beta chain, and derive the correct sequence of aminoacids. Then we use that sequence to synthesize the chain in our lab, because we have the correct information. Now, the result is the same in both cases. But, is semiosis implied? It is, certainly, in the first case. We derive the information from a symbolic representation, and we have to translate it according to a symbolic code that we already know. But not in the second case. The protein itself is not a representation of its own information. It is an instance of that information. We derive the digital sequence from the object itself. There is no symbolic code implied. Like in the example of the key. Proteins are like the key. One could ask: but is the sequence of AAs in the protein symbolic? Is it a code that stores information, for example, for its tertiary structure? And the answer is: no. The link between the linear sequence of AAs (the primary structure) and the final, tertiary structure, that generates the function, is defined by the laws of biochemistry. It is not symbolic. It is not a code. As you well know. That's why I insist that there are cases of functional digital information, that can well allow a design inference, which do not include semiosis and codes. It is absolutely true that most cases of digital functional information do include codes and semiosis. That is certainly true for language, mathematics, protein coding genes. IOWs, for most of the objects we are interested in. But it is equally true that there are cases of complex functional information without semiosis. As I have tried to show.gpuccio_{December 12, 2016
December
12
Dec
12
12
2016
10:53 AM
10
10
53
AM
PDT}

UB: I can agree with most of what you say. But the simple point is: for my procedure, I define digital information in a material object anything that can be read as a sequence of individual states that can represent an alphabet. In my example of the key, it is certainly possible to read the sequence of binary individual states as a digital binary sequence. Therefore, the sequence is digital information that can be retrieved from the object. And it has a connection with the function, because only is the sequence is the right one will the key work. In a sense, it is not different from any binary sequence that you can retrieve from the physical states of a CD surface. I have created this rather strange example to show that in exceptional cases digital information can be independent from a semiotic code. But I am not saying anything here about more abstract concepts, like representation. I am only saying that I can read a digital sequence that is connected to a function, and so I can apply my design detection procedure. Which is empirical, and does not depend on any theoretical considerations, other than the definitions given in the procedure itself.gpuccio_{December 12, 2016
December
12
Dec
12
12
2016
08:10 AM
8
08
10
AM
PDT}

8. dFSCI can only be found in mathematics and language No. It can be found in any form of digital information that implements complex functions. Software, or projects for machines. Of course, the digital nature usually requires some form of symbolic code, but that’s not necessary. Imagine a physical key which is made as a regular sequence of spikes and holes. It is digital, because it is a sequence of two possible states. If it is very long, it can be very complex. And it opens its safe only if the sequence is correct. But there is no symbolic code here: just a physical form that can implement a complex function, and can be described digitally as a sequence of two states.
Hello GP, good to talk to you. You have always been generally supportive of semiotic perspectives, even though I think we both know that we sometimes don’t see things in the same way (which is often a good thing). This is an interesting occasion. A lock on a safe is a device that controls access to the safe. Generally, such a lock has only a single variable – to ’allow access’ from a locked state. The shape of the correct key is a physical representation that communicates this single variable to the system. It accomplishes this be being measured at several specific points along the length of the key when it is pushed into the lock. In contrast to this, the central characteristic of a digital medium is that it is a sequence of many individual representations, which gives the digital medium its capacity to carry high levels of information. I think it’s fair to ask if a key being inserted into a lock is a set of individual representations or a single representation. If it is a set of individual representations, what do they represent? Do we take the position that each one represents some fraction of the total end effect? Do we also say that communicating a single variable is the “implementation of a complex function”? Let’s suppose that our key operates in a lock with 6 pins. Does this system represent “digital information” if the six pins are merely in a straight line, and therefore “not digital information” if the six pins are in a circle (such as in a tubular or radial key)? Let’s say we describe a 6-pin key (as you suggest) as “digital information” in the form of 101010. When I begin to push the key in, the first pin position will be at the “1” position and the other five pin positions will be empty. If I push the key in a little further, then the first pin position changes to a “0” and the second pin position becomes a “1”. Then they all change again when I push the key in further. It is only after I have pushed the key all the way in that the correct pattern can be recognized by the system. Is this a case of individual digits, or a single representation? I do not believe that a linear sequence of matter is sufficient for a representation to be considered a digital medium. It requires the establishment of digits. I am prepared to be convinced differently, and will amend my statement to incorporate the new understanding. Obviously, if we wish, we can say that a key contains digital information that is 1 digit long, but I am not certain how that clarifies anything.Upright BiPed_{December 12, 2016
December
12
Dec
12
12
2016
07:52 AM
7
07
52
AM
PDT}

@151 clarification:
Whether the biological systems are designed or not doesn’t depend on how much we know about it or how well we quantify, detect and infer design.
That statement was poorly written. What I meant was that even the areas of biology where we don't have any threshold-based quantification method to infer deign, still design can be inferred on the basis of other valid criteria. Basically the level of functional complexity* of the observed systems may be sufficient evidence to infer design. (*) functional coherence?Dionisio_{December 11, 2016
December
12
Dec
11
11
2016
01:53 PM
1
01
53
PM
PDT}

Silver Asiatic at #150: And thank you to you! :)gpuccio_{December 11, 2016
December
12
Dec
11
11
2016
12:54 PM
12
12
54
PM
PDT}

Silver Asiatic at #148: My answers: 1. All functions contain FSCI No. All functions contain FSI, because some information is needed to implement any function. But there is no general need that the information should be complex. In most cases, it is simple. 2. There is no difference in dFSCI between a Shakespeare sonnet and a poem that could have been generated by a computer If we consider dFSCI, IOWs, the binary form, there is no difference, because for both examples (provided they satisfy the conditions we have discussed before) the answer is: yes. Both examples exhibit dFSCI, and allow a design inference. But dFSI, in numeric form, is certainly different, if we use appropriate definitions. If we stick to a simpler definition satisfied by both, such as "being formed by english words", the dFSI linked to that function will be the same, because the function is the same and the length (I assume) is comparable. That's enough to infer design for both objects. But if we use some more refined functional definition, that is satisfied by the sonnet but not by the computer generated sequence, we will be able to compute higher dFSI for the sonnet, using that definition. However, the design inference remains valid for both objects. 3. Specification cannot be quantified It can be quantified as a binary variable (present - absent). if it is a functional specification (like, for an enzyme, being able to catalyze some reaction), then it is necessary to quantify the function and to have a threshold to assess if it is present or not (for example: the ability to catalyze the reaction at least at such level). That allows to get the binary value (present - absent) for any possible object. 4. ID is a probability measure False. ID is the theory of design inference. Of course, the theory includes some probability measures. 5. Cosmological fine tuning arguments cannot be analyzed for dFSCI Well, cosmological fine tuning arguments are about one object: the observable universe. Like biological arguments, they rely on the measurement of the information linked to our universe as an object able to implement some function (for example, being compatible with life) and what is believe to be the whole of possible universes (the search space). So, in a sense, the concept is the same: it is functional information here too, but applied to one object. Is it digital? That is a good question. Maybe we could ask some quantum physicist... 6. No microevolutionary events show evidence of dFSCI It depends on how we define "microevolutionary". If we mean simple transitions, of only a few bits (like most examples I am aware of), then the statement is true. There is some dFSI linked to the transition, but it is not complex (according to any reasonable threshold). 7. dFSCI is defined through subjective inputs on what a function is It is defined through an objective and shareable definition of a function, made by a conscious subjective being. Any function can be defined, by any observer. The dFSI is measured for the defined function. 8. dFSCI can only be found in mathematics and language No. It can be found in any form of digital information that implements complex functions. Software, or projects for machines. Of course, the digital nature usually requires some form of symbolic code, but that's not necessary. Imagine a physical key which is made as a regular sequence of spikes and holes. It is digital, because it is a sequence of two possible states. If it is very long, it can be very complex. And it opens its safe only if the sequence is correct. But there is no symbolic code here: just a physical form that can implement a complex function, and can be described digitally as a sequence of two states. 9. dFSCI analysis is not used, in itself, to reveal design, but only to validate what is claimed to be designed False. It is used to infer design for objects of which we don't know the origin (design or not). 10. Not all objects that contain functional specific information give evidence of design, but only those where the quantity of that information is high enough Absolutely true. We can infer design only if the functional information is high, IOWs if it is complex (above some appropriate threshold). OK, that was not difficult after all...gpuccio_{December 11, 2016
December
12
Dec
11
11
2016
12:48 PM
12
12
48
PM
PDT}

SA, Whether the biological systems are designed or not doesn't depend on how much we know about it or how well we quantify, detect and infer design. Generally objective truth does not depend on whether we know it. I understand that the quantification method gpuccio uses to digitally measure the functional complexity of certain biological objects and thus infer design is limited to those objects only. The brilliant ideas my former supervisor at work had could not be easily quantified. However, they were design ideas. The software we developed to implement those brilliant ideas could have been quantified using gpuccio's method to infer design. Different control layers and procedural components of the designed biological systems may or may not be suitable for quantification in order to infer design. Each of them should be analyzed using different methods. Perhaps some of those methods don't exist yet or may never exist. We are dealing with an unfathomable designed system that is beyond anything we conscious beings have ever imagined, much less designed. However, every day researchers from wet and dry labs are producing enormous amounts of new data that shed more light on the elaborate cellular and molecular choreographies orchestrated within the biological systems. We ain't seen nothing yet. The best is still ahead. It feels good to be on the winning side. But that also gives us the responsibility to be magnanimous toward those who disagree with us. Those of us who have been beneficiaries of Divine grace should enjoy being gracious toward others too! Let's enjoy learning from what serious science is discovering these days, specially in biology. Let's rejoice!Dionisio_{December 11, 2016
December
12
Dec
11
11
2016
12:00 PM
12
12
00
PM
PDT}

Dionisio 134 I feel the need to say this again. I agree with everybody here - even if you're mistaken. :-) Seriously, I am an ID advocate. Most importantly, Thank You -- to everybody, especially gpuccio! For taking so much time and painstaking detail to offer explanations. I hope more than just me benefited from it. A very good job done - again, thank you ALL IDers!Silver Asiatic_{December 11, 2016
December
12
Dec
11
11
2016
11:08 AM
11
11
08
AM
PDT}

UB I missed this earlier:
SA: For example, in origin of life studies, where there is no genetic code but just molecular activity. UB: This is an unsupported assumption, is it not?
I'm not sure here. If we're trying to evaluate the probability of non-living molecules coming together to form building blocks of life (pre-genetic code), then I'd think that would be right. But it's also true that we're looking for the origin of the genetic code, so yes - I understand and accept what you said.Silver Asiatic_{December 11, 2016
December
12
Dec
11
11
2016
11:04 AM
11
11
04
AM
PDT}

I'll close with a 10-question ID Quiz: True, False or Other (bad question, ambiguous, etc.) 1. All functions contain FSCI 2. There is no difference in dFSCI between a Shakespeare sonnet and a poem that could have been generated by a computer 3. Specification cannot be quantified 4. ID is a probability measure 5. Cosmological fine tuning arguments cannot be analyzed for dFSCI 6. No microevolutionary events show evidence of dFSCI 7. dFSCI is defined through subjective inputs on what a function is 8. dFSCI can only be found in mathematics and language 9. dFSCI analysis is not used, in itself, to reveal design, but only to validate what is claimed to be designed 10. Not all objects that contain functional specific information give evidence of design, but only those where the quantity of that information is high enoughSilver Asiatic_{December 11, 2016
December
12
Dec
11
11
2016
10:59 AM
10
10
59
AM
PDT}

UB
I think this statement is unfair, and perhaps a little bit opportunistic.
I apologize that I have gotten a little testy in my responses. I am a friend to ID, not an enemy -- and more importantly, an admirer of all who have contributed here and to UD for so many years. So again, I'm sorry that I came across that way. What I meant from "thin air" is that it sometimes appears that we're making it up as we go along. That's not a bad thing - perhaps we haven't been challenged on certain areas. I also don't think a difference of opinion among ID advocates is necessarily bad either. It's just good to air those things out. No, I didn't intend to mean that my ID friends and colleagues are being irrational or foolish or anything like that, and I apologize that it sounded that way.
I am not sure what “ID rules” are.
What I'm getting at is "how ID works" and "how we do or use calculations". These can be thought of as various rules. For example, as you said:
The clearly identifiable physical manifestation of dFSCI can be found in only three places anywhere in the cosmos — language, mathematics, and in the genetic code.
This could be considered a "rule". Why not? If everyone agrees, then this is one of the guidepoints of the ID process. "dFSCI can be found in only three places ..." But in making a rule, let's test it to make it better, if we can. That's all I was doing. So, I wondered if the genetic code was language. You stated:
In any case, if you analyze the genetic code from a physical perspective, it functions exactly like language,
I'd then say, "yes" the genetic code is a language, as other messaging codes are. If so, then the "rule" could be modified. Instead of "three places" we would say "only two".
The clearly identifiable physical manifestation of dFSCI can be found in only Two places anywhere in the cosmos — language and mathematics.
Right? I am very argumentative by nature and this is a character flaw most of the time. Part of that flaw is "wanting to win" whenever I get into a debate. I attempted here to try to "stir the pot" a bit and bring some questions. Let's face it, our opponents don't really do a very good job of this, most of the time. However, I don't like this position I've placed myself in. I notice my responses seem hostile, or as you righly said "unfair". I joined this blog years ago to make friends, not create enemies. :-) So, I see this little exercise I came up with to relieve boredom is not working well for that purpose. :-)Silver Asiatic_{December 11, 2016
December
12
Dec
11
11
2016
10:48 AM
10
10
48
AM
PDT}

Origenes
You seem to hold that, in order to be valid, the design inference should always work — also in difficult cases ...
I can understand why you think I'm questioning the validity of the inference - or more, that I think it's invalid. But I'm just looking for better explanations. I'm looking at what I think are weak points in the presentation. I never said that I thought the inference was invalid. I did say that I didn't understand many things. And I did point out a few conflicts among ID advocates on what various things mean. I consider myself a friend, not an enemy. I'll say also, I prefer your (our) individual responses rather than having to read papers or, worse, sit through youtube lectures. My goal - can we improve our own explanations? Can we explain what ID is and what the calculations are? Do we understand the limits? How precisely do we use these things? Have we tested the hardest cases? It's a challenge, not an attack. The better you can handle the hardest objections, in your own words, the more effective you (all of us) will be in explaining and defending.Silver Asiatic_{December 11, 2016
December
12
Dec
11
11
2016
10:33 AM
10
10
33
AM
PDT}

SA,
Well, I hope we’re having fun but it also seems like ID concepts and rules are being pulled from thin air – to add on top of the pile.
I think this statement is unfair, and perhaps a little bit opportunistic. As a general rule, I think you read people fairly carefully and thoughtfully (particularly the opponents of ID), which makes this statement all the more surprising. As for me, the “thin air” I am pulling from is directly out of the scientific record. As for GP, I would encourage you to read him more closely. If “thin air” is a euphemism for unwarranted or unprincipled reasoning, then you may find that the thin air is where the objections to dFCSI are coming from.
Speaking of ID rules … we might say that the genetic code is a language?
I am not sure what “ID rules” are. As far as I know, ID doesn’t have any rules that are not also part of any other empirical pursuit, (i.e. physical things must operate physically following the forces of physical law, with well-worn caveats to the unknown). In any case, if you analyze the genetic code from a physical perspective, it functions exactly like language, and indeed, the use of language is the only other physical process that the genetic code can be classified with.
But anyway, if there were strict limits to the use of dFSCI for understanding design in nature, then this should be known upfront before anyone tries to use it in other cases.
What to say? It just seems there is a conceptual block here that is causing unnecessary confusion.
For example, in origin of life studies, where there is no genetic code but just molecular activity.
This is an unsupported assumption, is it not?
I think that’s how we would have to do evolutionary analysis. Or as above, we would convert observations into statistics and then analyze the statistics and not the object.
I once saw a critic of ID on this blog say that we should be able to find out the “complexity” of a cake by writing out (in the least number of symbols) the process of making the cake, and then analyze the probability that such a sequence of symbols could come about. I never adopted that kind of thinking. My preference is to understand dFSCI at its physical embodiment. The confusion goes away quickly. cheersUpright BiPed_{December 11, 2016
December
12
Dec
11
11
2016
10:28 AM
10
10
28
AM
PDT}

Silver Asiatic, From your posts #113, #115 and #116 I got the notion that your concerned with, what I call, different starting points (#140). If I understand you correctly, you differentiate between easy starting points for design inference and difficult ones.
Silver Asiatic: Apply CSI, FSCI (whatever initials you want), to a variety of cases. No editorializing. No cover-ups. Just take a variety of randomly selected things. Apply the measure. Spell it out.

Silver Asiatic: ... that’s way too easy. It’s a human designed artifact with a known function (meaning in English).
You seem to hold that, in order to be valid, the design inference should always work — also in difficult cases:
Silver Asiatic:Yes, blind test with unknown languages. Test with languages that have partial function. Test with ambiguous function. Test with machine generated code, non-human designed- that has function. (Randomized parameters for evolutionary algorithms).
I find Dembski's explanation of the problem of false negatives very helpful:
Consider first the problem of false negatives. When the complexity-specification criterion fails to detect design in a thing, can we be sure that no intelligent cause underlies it? No, we cannot. To determine that something is not designed, this criterion is not reliable. False negatives are a problem for it. This problem of false negatives, however, is endemic to detecting intelligent causes. One difficulty is that intelligent causes can mimic necessity and chance, thereby rendering their actions indistinguishable from such unintelligent causes. A bottle of ink may fall off a cupboard and spill onto a sheet of paper. Alternatively, a human agent may deliberately take a bottle of ink and pour it over a sheet of paper. The resulting inkblot may look identical in both instances, but in the one case results by chance, in the other by design. Another difficulty is that detecting intelligent causes requires background knowledge on our part. It takes an intelligent cause to recognize an intelligent cause. But if we do not know enough, we will miss it. Consider a spy listening in on a communication channel whose messages are encrypted. Unless the spy knows how to break the cryptosystem used by the parties on whom she is eavesdropping (i.e., knows the cryptographic key), any messages passing the communication channel will be unintelligible and might in fact be meaningless. The problem of false negatives therefore arises either when an intelligent agent has acted (whether consciously or unconsciously) to conceal one's actions, or when an intelligent agent in trying to detect design has insufficient background knowledge to determine whether design actually is present. Detectives face this problem all the time. A detective confronted with a murder needs first to determine whether a murder has indeed been committed. If the murderer was clever and made it appear that the victim died by accident, then the detective will mistake the murder for an accident. So too, if the detective is stupid and misses certain obvious clues, the detective will mistake the murder for an accident. In doing so, the detective commits a false negative. Contrast this, however, with a detective facing a murderer intent on revenge and who wants to leave no doubt that the victim was intended to die. In that case the problem of false negatives is unlikely to arise. Intelligent causes can do things that unintelligent causes cannot and can make their actions evident. When for whatever reason an intelligent cause fails to make its actions evident, we may miss it. But when an intelligent cause succeeds in making its actions evident, we take notice. This is why false negatives do not invalidate the complexity-specification criterion. This criterion is fully capable of detecting intelligent causes intent on making their presence evident. Masters of stealth intent on concealing their actions may successfully evade the criterion. But masters of self-promotion bank on the complexity-specification criterion to make sure their intellectual property gets properly attributed. Indeed, intellectual property law would be impossible without this criterion. [source : 'No Free Lunch', p.24]
Origenes_{December 11, 2016
December
12
Dec
11
11
2016
09:48 AM
9
09
48
AM
PDT}

Silver Asiatic: You insist:
So, I concluded that there is no new dFSCI in Shakespeare because all his information can be found in the dictionary and rules of grammar. These are the same things programmed into the computer. The outputs are different though. One designed intentionally by a human author, the other was a non-conscious output from a machine. I wouldn’t equate both as “designed” in that case.
You conclude wrong. there is always dFSCI, in both cases. Both are designed. Of course, Shakespeare does mcuh more than simply filter random words by a dictionary or rules of grammar. But the design inference test gives the same result for all objects that have more than 500 bits of functional information according to any defined function. The two objects are different, but they are both designed.
Well I don’t think the computer generated poem is “designed”. It’s a non-consious output.
No. A non conscious output is an output generated in a non conscious system, a system where no conscious agent contributed to the output. That's not true here.
But it does seem you’re saying that there is no difference in the functional dFSCI information content of Shakespeare and the computer generated poem.
Not true. Of course there is higher functional information if Shakespeare's poem, but to show that we must use more refined function definitions, to show the specific meaning and creativity of the sonnet. But my purpose was not thins: it was to correctly infer design, and I have done exactly that: both objects are designed.
But this is a case of moving the goalposts. If additional specification is needed to determine if one output is designed or not, then the best answer to my test is that we cannot determine the difference in functional information between the artifacts.
You are making a very great confusion. As I have said tenths of times, both objects are designed, and both objects are recognized as designed. My example of the theorem was only made to show that there are things that a computer program cannot do algorithmically, but that anything can be apparently done if we input all the necessary information in a software. As an extrem example, if we input Hamlet in a software, the soaftware can output it, both as the result of a long filtering of random words, like in the "Methinks" case, or simply by printing the information it already has. That has nothing to do with the problem of design inference: as I have said tenths of times, I have inferred design for both objects you proposed, and I do believe that they are both designed, and that my two inferences are two true positives. You are free to disagree.gpuccio_{December 11, 2016
December
12
Dec
11
11
2016
08:37 AM
8
08
37
AM
PDT}

Silver Asiatic:
That may be fine, but you know nothing of that when you observe the artifacts I presented. As I mentioned to Origenes above, I gave away the answer to the problem first. In all three cases, you would find the same level of “conscious output” but in fact, the designer of the computer poem was not conscious of the output. To say that the computer poem was ‘designed’ is to say that any random computer program shows evidence of design, because there is some order or use of regularity.
No. My evaluation of your examples did not depend on the fact that you had "gave away the answer". Please, look at my post #132, where I comment on your examples. As you can see, my answer is clear and unequivocal: All your examples will be considered designed according to may procedures. And, according to the information you give about their origin, all your examples are designed. there fore, they are all true positives. You say that "the designer of the computer poem was not conscious of the output." I absolutely disagree. Please, read again what I have already written: "OK, but let’s assume that a very complex computer, with a very complex software and a lot of information about grammar implemented, can generate a “poem” that is grammatically correct. I can only repeat what I already said. The poem can be recognized as designed, because of the information about words and grammar that it exhibits. If we have just the poem, we can safely infer that it is designed. And we are right. It is a true positive. But again, all the functional information that we are observing comes from the programmer of the software. He is the conscious agent who represented and implemented that information (sequences that respect grammar, generated by the machine I am implementing, acting on random seeds). Again, the random component adds no functional information to the output. The functional information we observed is designed. By the programmer." You say: "To say that the computer poem was ‘designed’ is to say that any random computer program shows evidence of design, because there is some order or use of regularity." I really cannot find any sense in that. What is a "rnadom computer program"? What kind of "order and use of regularity are you referring to? Why should that supposed "order" be "evidence of design"? Please. read again what I have already written: "The computer created exactly what the programmer predicted: a sequence that respects the rules of grammar. He predicted it, and he obtained the desired result. He certainly did not predict the specific random components of that sequence, but he certainly predicted that there would be a contingent component of the output that he could not anticipate. He certainly predicted that such a contingent component would be, indeed, contingent, random, and would add no further functional information to the result. Again, all the functional information we observe in the result is from the programmer." So, there is no problem for me: all your examples are designed, and all your examples are correctly categorized as designed applying my procedure. If you have problems, you should explain them better. If you simply disagree, you are free to disagree. I certainly disagree with you. You ask:
Yes, but how do we reach that conclusion simply by looking at the various outputs?
You seem to expect that I have reached some conclusion about the difference between a poem generated by a programmer through a software and a poem written directly by a poet. But I said nothing about that. For me, both appear designed according to a dFSCI analysis. For me, both are designed according to what you say of their origin. The design inference must distinguish between designed objects and non designed objects. Both your examples are designed, and both are correctly categorized as designed by my procedure. I have done nothing to distinguish between them. As a more general point, I have briefly discussed some of the limitations of the output of computers. But I have never said that my procedure can distinguish between the output of a computer and the direct output of a human poet. That is probably possible, but it would not be a problem of design detection, but rather of design differentiation: distinguishing between two different modalities of design. But I have said nothing about that. My interest is to detect design, not to differentiate between between design by pen and paper and design through a computer. Moreover, I have clearly explained that a computer can generate a completely detailed and creative design, even a full poem, if the creative information is pre-loaded in the software. This concept is a very simple expression of the law of conservation of information. And it is true. Computers cannot generate new original functional information, but they can certainly use the information they have received. To Origenes you say:
Well, gpuccio would say that the computer poem was designed. I would call that a false positive. To say that “the programmer created the randomization patterns that resulted in the poem”, is really to say that the output of every computer randomization routine (including a random character generator) is “designed”.
But I never said that the programmer "created the randomization patterns that resulted in the poem". What do you mean by "randomization pattern"? The programmer does two things: 1) He generates random sequences (which, by definition, are generated according to no pattern, because they are randomly generated"). 2) He filters those sequences according to a dictionary (and then to grammar rules). Of course, he inputs in the software tons of functional information to do that (do you know how many bits are only in the dictionary?) That filtering and those rules generate the "patterns" that are recognized as functional in the dFSCI analysis. That is exactly what the programmer envisioned and implemented. That is, without any doubt, intelligent design. And as intelligent design it is recognized by the dFSCI procedure. So, no false positives at all. A true positive. More in next post.gpuccio_{December 11, 2016
December
12
Dec
11
11
2016
08:21 AM
8
08
21
AM
PDT}

UB @136:
The clearly identifiable physical manifestation of dFSCI can be found in only three places anywhere in the cosmos — language, mathematics, and in the genetic code.
Interesting statement. Thank you.Dionisio_{December 11, 2016
December
12
Dec
11
11
2016
07:07 AM
7
07
07
AM
PDT}

Silver Asiatic, The starting point, or starting level, of a design inference seems quite arbitrary to me. One has to ‘give’ the materialist something to work with. That may be ‘monkeys and typewriters’ or a computer program with inbuilt dictionary and grammar rules — one has to start somewhere. However some starting points make the design inference very difficult, for various reasons.
Silver Asiatic: 1. How did dFSCI analysis show that Shakespeare just merely the dictionary and rules of grammar and poetics? Just as the computer did? 2. How much, precisely, new information did Shakespeare’s poem contain over the computer generated poem?
This may very well be a problem. If so, then the problem is due to the starting point. IOWs we have to choose the starting point for the design inference wisely. GPuccio’s homology based argument wrt proteins (see e.g. #82) does exactly that. It starts out with organisms capable of creating proteins. So, what GPuccio does here is saying: ok, let’s start at this level and suppose that everything up till this level can come into existence by natural means. Now, let’s see if we can explain the coming into existence of observed new information from here on. IOWs starting at this point, is extremely generous towards materialism. It momentarily brushes aside a host of design arguments — including bio-semiose and fine-tuning arguments. But, in spite of all this generosity, design is easily proved. So this is an excellent starting point! An important property of GPuccio’s argument is that ‘function’ can be easily defined and agreed upon. Naturalism is grappling with the concept 'function', but here not so much:
Accounts of biological function which refer to natural selection typically have the form that a trait's function or functions causally explain the existence or maintenance of that trait in a given population via the mechanism of natural selection. [ Stanford website]
GPuccio’s argument fully utilizes this naturalized definition of function.Origenes_{December 11, 2016
December
12
Dec
11
11
2016
06:53 AM
6
06
53
AM
PDT}

gpuccio
Now I will explain why I don’t believe that a computer program can create a poem that is originally creative. Th simple idea is: non conscious system cannot create new specifications (meanings or functions that have not been already programmed in them) for the simple reason that they do not understand what meaning and purpose is: they have no subjective experience.
That may be fine, but you know nothing of that when you observe the artifacts I presented. As I mentioned to Origenes above, I gave away the answer to the problem first. In all three cases, you would find the same level of "conscious output" but in fact, the designer of the computer poem was not conscious of the output. To say that the computer poem was 'designed' is to say that any random computer program shows evidence of design, because there is some order or use of regularity.
IOWs, they are machines, and nothing else.
Yes, but how do we reach that conclusion simply by looking at the various outputs?
SA: However, you’re saying that since Shakespeare used information from the dictionary, Shakespeare is only recycling information from the dictionary, and there is no new information presented, right? The sonnet has no dFSCI? GP: I don’t know why you have this strange idea. I am not saying that, and I have never said anything like that.
Well, I'm looking for the difference between Shakespeare and the computer poem. How do you know what a human author did to create the poem. How is that observed? I will change the quote: “If a Shakespeare generates a sequence formed by correct english words by filtering a random output according to a dictionary, the information in the output comes from the dictionary, not from the random output. Shakespeare is only recycling the information it already contains (the dictionary), using also the computational procedures programmed in Shakespeare's brain. There is no generation of new original dFSCI.” So, I concluded that there is no new dFSCI in Shakespeare because all his information can be found in the dictionary and rules of grammar. These are the same things programmed into the computer. The outputs are different though. One designed intentionally by a human author, the other was a non-conscious output from a machine. I wouldn't equate both as "designed" in that case.
What I am saying is that the information in a dictionary is designed, and the information in the software is designed. Therefore, if we infer design for a sequence generated that way (made of correct English words), and we can, according to my procedure, we are correct: that sequence has been designed, indirectly, by the programmer who wrote the software and implemented it with a dictionary.
Well I don't think the computer generated poem is "designed". It's a non-consious output. But it does seem you're saying that there is no difference in the functional dFSCI information content of Shakespeare and the computer generated poem.
Indeed, the function we define is “being formed by correct english words”. That function was conceived, specified and implemented by the designer of the software. We recognize it and correctly infer that it was designed.
Yes, but the actual poem was not designed.
The important point is: the non conscious system formed by the software and the dictionary did not add any further complex functional information: its only contribution is the contingency of what specific random words are in the sequence, and that contingency is random, and includes no further meaning or complex functional information.
How do we know that difference between a poem written by a human author (which you are observing side by side with a computer generated poem) -- without knowing how the author wrote the poem?
Therefore, the complex functional information we observe in the object was all designed by the programmer. And our inference of design is perfectly correct: it is a true positive.
The complex functional information would not then, be linked to the actual content of the poem, but only to the structural regularities (grammar, syntax). As I said above, this would yeild a false positive if the task was to determine which specific output was designed.
And here the important point becomes clear, like a shining sun: can a software programmer write a software that will output a sequence made of correct english words, that respect the rules of grammar, and that conveys the demonstration of Pytagoras theorem, starting from randomly generated sequences?
But this is a case of moving the goalposts. If additional specification is needed to determine if one output is designed or not, then the best answer to my test is that we cannot determine the difference in functional information between the artifacts. Shakespeare, human poem and computer poem all have the same functional information content. But I would also consider that a false positive in design analysis. The computer poem is not designed, unless we're going to say that the output of mutation and selection is evidence of design also since without physical laws, chemical regularities, then no biological objects could exist.Silver Asiatic_{December 11, 2016
December
12
Dec
11
11
2016
06:31 AM
6
06
31
AM
PDT}

Origenes
However at the level of sentences (and higher levels) Shakespeare creates new information.
This is good and I will repeat this in my response to gpuccio, but the problem here is: 1. How did dFSCI analysis show that Shakespeare just merely the dictionary and rules of grammar and poetics? Just as the computer did? 2. How much, precisely, new information did Shakespeare's poem contain over the computer generated poem?
What seems to be important here is the notion that the design inference is on one end (many false negatives) an imprecise instrument, but yet can detect design very reliable (no false positives).
Well, gpuccio would say that the computer poem was designed. I would call that a false positive. To say that "the programmer created the randomization patterns that resulted in the poem", is really to say that the output of every computer randomization routine (including a random character generator) is "designed". But this doesn't show the key difference. It would be saying that natural forces (rules and grammar) of nature combined with randomization (mutations) create something designed.
The difference is that a computer has no intention, plan, meaning, teleology — whatever the appropriate term is. When we can formulate a specification wrt to a poem, mechanism, process or object we can ‘retrieve’ the intention, plan, meaning or teleology of the designer. The fact that a specification is possible provides us with an argument for the idea that teleology has occurred. Teleology, in turn, points to an intelligent designer.
I find this to be very good, but as UB put it, you seem to be tossing an entirely new criteria on to the pile. How does our analysis of dFSCI indicate the intent found in the various poems? I fully agree, that the computer has no intent (purpose, meaning). Additionally, I will say that the computer is "not conscious". It exhibits and intelligent output (gpuccio disagrees), but how would we determine the difference in terms of information, function and teleology between that and a similar poem. Keep in mind, computer generated poetry is already pretty sophisticated. It is being used in various applications to generate new ideas, new word combinations. True, nobody is publishing computer poems for the value they have in themselves. But I will repeat this to gpuccio ... You guys are kind of 'cheating' here. I gave away the secret (the final answer) to the test. And it seems you all jumped on it. I told you upfront, that one of the poems was computer generated. And it seems like you analyzed that poem, already knowing its source. Yes, but the harder part is to do the analysis without knowing which was human and which computer generated. Then match the results to the analysis on Shakespeare also.Silver Asiatic_{December 11, 2016
December
12
Dec
11
11
2016
06:02 AM
6
06
02
AM
PDT}

UB
As long as everyone is throwing their agreements and disagreements in a pile ???? ,
Well, I hope we're having fun but it also seems like ID concepts and rules are being pulled from thin air - to add on top of the pile.
I’ll contribute by saying that whole pieces of this conversation have baffled me.
] I'm more baffled now than before we started - but that's not necessarily a bad thing.
The clearly identifiable physical manifestation of dFSCI can be found in only three places anywhere in the cosmos — language, mathematics, and in the genetic code.
Speaking of ID rules ... we might say that the genetic code is a language? But anyway, if there were strict limits to the use of dFSCI for understanding design in nature, then this should be known upfront before anyone tries to use it in other cases. For example, in origin of life studies, where there is no genetic code but just molecular activity. Or in the cosmos itself, with fine-tuning. Now we could say for that, that we convert the observations we see into mathematics but see below.
It also seems odd to describe some object in language, and then suggest that your description reflects the dFSCI in the object.
I think that's how we would have to do evolutionary analysis. Or as above, we would convert observations into statistics and then analyze the statistics and not the object. my $.01 -- if indeed that much. :-)Silver Asiatic_{December 11, 2016
December
12
Dec
11
11
2016
05:47 AM
5
05
47
AM
PDT}

As long as everyone is throwing their agreements and disagreements in a pile :) , I'll contribute by saying that whole pieces of this conversation have baffled me. The clearly identifiable physical manifestation of dFSCI can be found in only three places anywhere in the cosmos -- language, mathematics, and in the genetic code. It seems pointless to me to ask if a mountain or a snowflake has any dFSCI. It also seems odd to describe some object in language, and then suggest that your description reflects the dFSCI in the object. my $.02Upright BiPed_{December 11, 2016
December
12
Dec
11
11
2016
04:14 AM
4
04
14
AM
PDT}

Origenes, Dionisio: Thank you for your comments, I agree with what you say, and I agree that this discussion is very improtant, whatever the individual positions. I invite all to look at Robert Mark's video here: https://uncommondescent.com/intelligent-design/prof-bob-marks-on-what-computers-cant-do/ which gives many important insights, and, for those really interested in the problems with strong AI, I would also definitely suggest reading this precious paper by johhnyb (who is also the author of this OP): http://www.blythinstitute.org/images/data/attachments/0000/0041/bartlett1.pdf Just a final hint, that is IMO very important: ID theory is a new paradigm that falsifies not only neo darwinian theories in biology, but also so called "strong AI theories" (in the sense given by Penrose) and the theories of consciousness related to those AI theories. Such is its importance!gpuccio_{December 11, 2016
December
12
Dec
11
11
2016
04:02 AM
4
04
02
AM
PDT}

Prev 1 2 3 4 5 … 8 Next

You must be logged in to post a comment.

Leave a Reply