|
Author
|
Topic: Towards a simple definition of CSI
|
Irving
Member
Member # 535
|
posted 14. March 2006 19:49
Bruce,
quote:
I fail to see how any thinking person can reject the general definition of specified complexity: complex (having limited compressability even knowing the nature of the data) information which is used to generate a "thing" which is capable of performing intricate tasks.
Perhaps I'm not a thinking person.... ![[Smile]](smile.gif)
I see Specfified Complexity as an adjective that describes a particular type of information. As a kid, I had a toy called a spirograph which was pretty simple, but created some very intricate things.
quote:
I propose that "intelligence" be defined as the process of generating (creating) information.
quote:
However, if Dembski suggests CSI as proof of intelligent design, I must challenge that by a reasonable definition of intelligence RM+NS is an agent of intelligence in and of itself.
Well, that's the problem then. If you define Intelligence that way, then there isn't really any difference between Intelligence and Nature.
I would suggest Intelligence as: That which is capable of abstract reasoning.
IP: Logged
|
|
David L. Hagen
Member
Member # 323
|
posted 14. March 2006 21:37
Bruce and Doug
As we are discussing the origin of information, can we agree on foundations or presuppositions?
See:
Reverse Engineering Assumptions for an Open Science Intelligent Design Theory
In particular:
quote:
5) Openness: Observable phenomena may be within open systems accessible to the input or intervention of intelligent causes, some of which may be detectable, and might be reproducible.
(E.g., the use of remote observation and communications systems by casino players and their accomplices that result in actions noticeably different from stochastic processes. By doing so, we assume the historic presuppositions by almost all founders of modern science, including modeling systems as “black boxes” or "gray boxes" where there may be causes and phenomena that we do not yet know of or fully understand. In particular, while assuming order and stochastic processes in the universe attributable to natural laws, we relax any a priori presumption that science can only address closed systems operating under known laws. We explicitly reject any exclusive "materialistic naturalism," and allow that there may be other sources for observed phenomena.
If we are to compare theories, we need to have an equal basis for evaluating them, and thus must start with the option that both Intelligent Design and Darwinian evolution (RM/NS) are to be compared.
This requires that we assume:
1)Open science: Assume that the universe is open with the real possibility of an intelligent designer behind biotic information as well as natural causes of the four forces, and that such cause may be detectable through objective evidence.
By contrast if you require that we ONLY use:
2) Closed science:
Assume a closed universe with only natural causes, a priori excluding intelligent causes.
Then we cannot address or compare origin theories of intelligent vs naturalistic causes for information. i.e., closed science requires than no designer exists, or that a designer can have no input into the universe, or that a prioriwe refuse to address any theories involving a desiger, whether there is testable objective evidence or not. Then since life exists, the logical consquence is that biotic information "must" have come from natural causes and thus by RM/NS.
Can we agree on an Open Science? If not, most of this discussion will be a futile effort in miss-communication because ID is dismissed a priori.
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 15. March 2006 12:51
quote:
Bruce,
quote:
I fail to see how any thinking person can reject the general definition of specified complexity: complex (having limited compressability even knowing the nature of the data) information which is used to generate a "thing" which is capable of performing intricate tasks.
Perhaps I'm not a thinking person....
I see Specfified Complexity as an adjective that describes a particular type of information.
You don't seem to be debating "complex (having limited compressability even knowing the nature of the data) information which is used to generate a "thing" which is capable of performing intricate tasks." doesn't exist, but rather you seem to be debating whether that description rightly fits "CSI".
Let me introduce you to a term, "overloading". (a term taken directly from Object Oriented programming.) In language, we use "overloading" all of the time, which is to say, we apply multiple definitions to the same term. This is why dictionaries often present multiple definitions, sometimes with two primary entries, and sometimes with numbers within the definition.
I am happy to say that I have defined something as "CSI" with a slightly different defintion than Dembski has defined it. Albiet, not that different, as I am attempting to achieve a summary definition of his work. I think you will find that my definition applies to computer sofware, to DNA and only to complexities of that ilk.
Further, please feel free to show where my definition of CSI is substantively different than Dembski's.
quote:
quote:
I propose that "intelligence" be defined as the process of generating (creating) information.
quote:
However, if Dembski suggests CSI as proof of intelligent design, I must challenge that by a reasonable definition of intelligence RM+NS is an agent of intelligence in and of itself.
Well, that's the problem then. If you define Intelligence that way, then there isn't really any difference between Intelligence and Nature.
I would suggest Intelligence as: That which is capable of abstract reasoning.
Your input is the exact kind of frustration that I have in these discussions. There is far too much trying to achieve the end from the beginning, rather than observing every step.
We cannot use "Dembski's CSI = intelligence" as a premise when we are working the proof that Dembski's CSI = intelligence.
A question that must be addressed, for starters is, What is information, how is it created? My conjecture is that any "order" is information, and anything that produces "order" creates information. If I run gravel through a screen, I get a tiny bit of order, that gravel which went throught the screen is smaller than a certain size. This appears to me to be information -- not Complex, Specified information, but information. A filter, therefore, "creates" information which was not there before.
If you choose to define "intelligence" as "That which is capable of abstract reasoning" you will first have to provide a good definition of "abstract reasoning", then you will have to provide a proof that CSI can only be created via abstract reasoning before you can claim CSI = intelligence.
As far as I understand it, that's the way that logical proof works. I personally am unconvinced that abstract reasoning is required to produce CSI.
David L. Hagen,
I have been reading your "Open" thread with interest. I am open to "open". I think that inn 1998 when "dark energy" was discovered, the assumption that the universe is a closed system was fundimentally questioned.
IP: Logged
|
|
Irving
Member
Member # 535
|
posted 15. March 2006 15:38
Bruce,
quote:
This appears to me to be information -- not Complex, Specified information, but information. A filter, therefore, "creates" information which was not there before
I don't think the problem will be with the definition of Information. The problem is will be with pinning down a definition of Specification.
quote:
If you choose to define "intelligence" as "That which is capable of abstract reasoning" you will first have to provide a good definition of "abstract reasoning", then you will have to provide a proof that CSI can only be created via abstract reasoning before you can claim CSI = intelligence.
Yes, and that is what Dembski has attempted to do with Specified Complexity. Given that we can't use knowledge of the Designer, all we are left with is knowledge of Nature. Thus the definition of Specification is Abstract Reasoning. Abstract Reasoning, is Independent, Detachable reasoning. Nature deals with the here & now...the local environmental conditions & laws. Abstract reasoning can leap beyond the here & now, and conceptualize the far away, and future/past.
This is why mathematics is so often used when discussing Specified complexity. The "concept" of mathematics is an abstraction from the physical world. The "concept" of prime numbers is independent & detachable from the physical properties that result in the manifestation of radio waves.
A sequence of prime numbers is complex information. Though highly improbable, a random number generator might just produce a sequence tomorrow, or not in the next 100 billion years... Who knows if heads or tails will be next. But the sequence of prime numbers is not Complex Specified Information until it is manifested in physical reality by process that are independent of such conceptualization.
quote:
What is specified complexity? Recall the novel Contact by Carl Sagan.16 In that novel, radio astronomers discover a long sequence of prime numbers from outer space. Because the sequence is long, it is complex. Moreover, because the sequence is mathematically significant, it can be characterized independently of the physical processes that bring it about. As a consequence, it is also specified. Thus, when the radio astronomers in Contact observe specified complexity in this sequence of numbers, they have convincing evidence of extraterrestrial intelligence. Granted, real-life SETI researchers have thus far failed to detect designed signals from outer space. The point to note, however, is that Sagan based the SETI researchers’ methods of design detection on
actual scientific practice.
From: Dembski's Expert Witness Report
Note: "...characterized independently of the physical processes that bring it about."
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 15. March 2006 16:00
quote:
Thus the definition of Specification is Abstract Reasoning.
If this is the definition of specification, then the knee-jerk response from the scientific community must be to reject it.
For there to be a valid definition of CSI, that definition must not force the fregone conclusion that CSI = intelligence.
The purpose of this thread is to come up with such a defintion. The latest posts clarify what I see as the problem, the assumption that CSI = intelligence.
It would seem to me that the definition of CSI must encapsulate that "the order of the nuclotides in DNA is a big deal" factor, it must also include "intelligently made" computer software. It must, however, disclude stuff such as crystaline structures, weather patterns, and all manner of other stuff that has very complex natures but doesn't have that "je ne sais quoi" characteristic of DNA and computer software.
Let me propose that we are seeking a definition (possibly an overloaded definition) for a term we will call "specification". This term means the kind of complex order / importance of order as found in DNA, and computer programs which differentiates from the kind of complexity found in crystals and weather patterns.
Let me also request, that the a priori concept of "intelligence" (defined as "which is capable of abstract reasoning") cannot be included in that definition.
Only once we have such a definition can we possibly determine whether the only reasonable method of getting such specified complexity is via "which is capable of abstract reasoning" intelligence.
IP: Logged
|
|
Irving
Member
Member # 535
|
posted 15. March 2006 21:26
quote:
Let me propose that we are seeking a definition (possibly an overloaded definition) for a term we will call "specification". This term means the kind of complex order / importance of order as found in DNA, and computer programs which differentiates from the kind of complexity found in crystals and weather patterns
Perhaps your putting the cart before the horse, or maybe not recognizing that there is a cart AND a horse. Or not, it might just be me...
Consider this quote from Explaining Specified Complexity
quote:
Does nature exhibit actual specified complexity? This is the million dollar question. Michael Behe's notion of irreducible complexity is purported to be a case of actual specified complexity and to be exhibited in real biochemical systems (cf. his book Darwin's Black Box). If such systems are, as Behe claims, highly improbable and thus genuinely complex with respect to the Darwinian mechanism of mutation and natural selection and if they are specified in virtue of their highly specific function (Behe looks to such systems as the bacterial flagellum), then a door is reopened for design in science that has been closed for well over a century. Does nature exhibit actual specified complexity? The jury is still out.
Dembski's purpose is/was to determine why humans appear to have an ability to differentiate Design from Nature. Or perhaps better said...What characteristics are there about Designed objects that differentiate them from those produced by Nature. Dembski proposed that Specified Complexity IS that characteristic.
Whether life exhibts CSI is a secondary question.
If Nature can produce CSI, then the whole endeavor provides no purpose. Thus as you note, the definition of Specified Complexity has been carefully crafted to exclude Nature. So in accordance with Dembski's first proposition:
Has the definition of Specified Complexity been established such that Nature cannot produce it?
If YES, then (and only then), is it fruitful to see if Life or any other objects exhibit Specified Complexity.
quote:
Only once we have such a definition can we possibly determine whether the only reasonable method of getting such specified complexity is via "which is capable of abstract reasoning" intelligence.
Dembski offers a defintion which defines Specification as an Independent, detachable pattern.
Yet, since that definition seems to rule out Nature you don't like it?
quote:
Further to the lack of a concise definition of CSI, the feeling that the thread left me with was that any definition of CSI was carefully loaded to make sure that RM+NS could not possibly produce a product that meets the definition.
Which, as I've described is the whole point.
Besides, the possibility that defining Specification as an independent, detachable pattern eliminates RM + NS...what do you see is wrong with it?
IP: Logged
|
|
Christopher D. Beling
Member
Member # 723
|
posted 15. March 2006 21:32
quote:
the definition of Specification is Abstract Reasoning. Abstract Reasoning, is Independent, Detachable reasoning.
This may be true but there is a better way to go about defining Specification. One needs an objective definition - not just because such can stand in the scientific community - but also it is a helpful benchmark. Such a definition is available. It is very simple and straightforward - it is to be defined through FUNCTIONABILITY (which equates with SURVIVABILITY - or REPLICATIONABILITY ).
Take a mathematical space X, [In terms of CSI we may take this to be a complex space]. You might think of X as being the configuration on a strand of DNA, or perhaps the set of numbers in an engineering office - numbers that have the potentiality of when being implemented in some factory of being able to make FUNCTIONAL machines like airplanes and computers. Most of the elements of X will DO exactly nothing (i.e. non working machinary of just a load of junk pieces). This may be expressed mathematically as follows:
S (X*)=SxX*
That is X* has the quality of SPECIFICATION it is an eigenstate (eigenvalue) of the operator S. What is this operator? It is a mathematical operator (exceedingly complex and embodying physical law) that checks all possible system configuration states X to see if they are in some sense FUNCTIONAL=SURVIVABLE when implemented. Experimentally S is more simple - just energize and see if the system SURVIVES (REPLICATES). Take for example a long piece of DNA that codes for a Giraffe - this X produces a functional system X* - the Giraffe that can SELF PROPOGATE (in conjunction with other Giraffes!). The same would apply to airplanes. If an airplane flies well then manufactures will want to make more of it - i.e. the design survives. However if it does not fly then S=0 and no manufacturer is going to make it!
Thus
if S=0 -> not specified
if S>0 -> specified
To understand that operator S more succinctly we may call it also the SURVIVABILITY operator. If when operating on a realized configuration X* it brings back the same system X* (next generation), then it has satisfied the SURVIVABILITY (or replicationability) test. The system is then classified as a system that is SPECIFIED, and the code X* that leads to the system as having SPECIFICATION.
Note: This definition is a SUFFICIENT condition for SPECIFICATION but NOT A NECESSARY condition. For example the set X of prime numbers is considered by many to be a specification - because it needs intelligence to produce this pattern. It is for practical consideration only necessary to have the above definition of specification - for if this condition is met - for certain the configuration X* (or system) is specified.
Hope you agree: Chris
[ 15. March 2006, 22:11: Message edited by: Christopher D. Beling ]
IP: Logged
|
|
Irving
Member
Member # 535
|
posted 15. March 2006 22:02
Chris,
The definiton of Specification is straight forward as defined here at ISCID in the encyclopedia.
I don't believe Dembski is using the term Specification in a strict engineering sense. He's "overloading" the term to use Bruce's terminology.
quote:
For example the set X of prime numbers is by considered by many to be a specification - because it needs intelligence to produce this pattern.
Why does it require Intelligence? Because it is Abstract from the Physical World.
quote:
...it is to be defined through FUNCTIONABILITY (which equates with SURVIVABILITY - or REPLICATIONALITY ).
Why, again? I don't consider SURVIVABILITY as necessary to the defintion of CSI.
IP: Logged
|
|
David L. Hagen
Member
Member # 323
|
posted 15. March 2006 23:51
Thanks Bruce for that affirmation on the assumptions for Open Science.
1) What is CSI? The perceived consequences of Human Intelligence vs abiotic Universe
I think what we are trying to quantify is what do we as intelligent beings do that we perceive is different from abiotic nature?
e.g., What is it that hardware or software engineers design, or inventors patent?
2) Prepare Origin Free Definition of CSI
Then define this without embedding either natural abiogenesis (popularly "evolution") or an Intelligent cause? e.g. can we define biotic life without embedding natural abiogenesis or Intelligent Cause?
3) Test Origin Models vs Definition & Biotic Evidence
Then model and quantify to see if CSI that can be obtained by natural abiogenesis (from abiotic nature in a closed system of natural causes. e.g, the four forces within a closed universe.)
4) Inferrence
If natura abiobenesis cannot meet the UPB, then the inference is that an Intelligent Cause is most probable as we see in every example where human beings are the causes of intelligent activity, and as modeled and searched for in SETI.
I thought that I was doing this in my previous posts - not that I was embedding intelligence as the only origin or cause into the definition of CSI. Distinguish by challenging the feasibility of natural abiogenesis with embedding an intelligent cause into CSI.
5) CSI Definition: quote:
In precise form, CSI = the property of their being some event or object J, with complexity K conforming to an independent pattern L.
K must meet some high threshold, normally the UPB.
Does Micah's precise definition require or imbed either ID or Evolution as necessary origin theory?
Note that Dembski's "4th Law" is defined in the negative with no mention of an intelligent cause.
6) Self Reproducing Life
This CSI in its broadest sense does not obviously include "FUNCTIONABILITY (which equates with SURVIVABILITY - or REPLICATIONALITY )".
However, when defining the simplest biotic life, I expect that the capability for self reproduction or equivalent will enter the definition. This in itself does not officially embed either ID or natural abiogenesis (popularly claimed for evolution by RM/NS).
After modeling, natural abiogenesis as an origin theory may be found to not be a possible cause within the Universal Probability Bound (UPB), leaving ID as the only credible alternative. ID further has the task of forming comprehensive, coherent, practical, and predictive theories.
Note that Darwin and Yockey both presume that the origin of life exists prior to evolution by RM/NS. I.e., by definition, natural selection cannot happen until we have self-reproducing life. I do not know of any natural theory of abiogenesis that comes close to claiming the origin of life with a probability greater than the UPB.
[ 16. March 2006, 11:44: Message edited by: David L. Hagen ]
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 16. March 2006 12:24
quote:
Dembski offers a defintion which defines Specification as an Independent, detachable pattern.
I personally like this defintion. Specification is a detatchable pattern, if the pattern is moved into a new machine that actualizes that class of patterns, the result will be the result defined by the pattern.
If we take the "independant detatchable pattern" for a human ear, and stick it into a mouse, the mouse grows an ear -- out of it's back. If I take a detatchable pattern for a word-processing program and stick it into a computer, the computer does the word-processing function. This is the thing that separates (DNA, computer programs, etc.) from (crystals, weather, etc.) This definition would disclude all abiotic nature, and would include virtually all human technology (as virtually all human technology has an independant detatchable specification (blueprint, flowchart etc.))
quote:
Yet, since that definition seems to rule out Nature you don't like it?
I don't know where you get this. I am not seeking a definition that "does not rule out nature", I am only seeking a definition that fits the data rather than fitting a philosophy. If, by time you have a definition that fits the data, some philosophy is the only viable explanation, well, so be it.
This definition however: "specification = an independant, detachable pattern" does not seem to justify your quote from Dembski,
quote:
Does nature exhibit actual specified complexity? ... Michael Behe's notion of irreducible complexity is purported to be a case of actual specified complexity ... Does nature exhibit actual specified complexity? The jury is still out.
The above definition seems to apply to all coding DNA, all DNA but true junk DNA. I fail to understand, then, why Dembski would suggest that Behe's "IC" concept is required to establish specified complexity.
IP: Logged
|
|
David L. Hagen
Member
Member # 323
|
posted 16. March 2006 16:07
Bruce
Those are good examples of "independant detatchable pattern." Now to model the simplest cell and challenge if that can be arrived by abiogenesis and quantify the probability vs the UPB.
-------------
quote:
why Dembski would suggest that Behe's "IC" concept is required to establish specified complexity.
I think Dembski actually said the opposite. i.e. that Behe's IC may be an example of specified complexity, not required for it. Please reread Dembski:
quote:
Michael Behe's notion of irreducible complexity is purported to be a case of actual specified complexity
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 16. March 2006 16:31
It still feels to me that Dembski's definition of CSI is rather more restricted than the definition that is being worked here:
quote:
CSI is any packet of information containing: information (order) with Specification (an independant detatchable pattern) and sufficent complexity (having limited compressability even knowing the nature of the data) to extend beyond UPB.
However, I believe that a definition like that above -- a definition which includes coding DNA, and human-generated instructions but rejects (virtually) all of the abiotic world is called for.
I do not believe that the scientific community would reject such a definition. (Or if they do it would be because of a haunting fear that it is some sort of wedge. The latter I see as gutlessness, though I recognize it as a phenomenon amongst both parties in virtually all debates.)
We can now get to the meat questions such as David Hagen's quote:
Now to model the simplest cell and challenge if that can be arrived by abiogenesis and quantify the probability vs the UPB.
I have not found any way to get to CSI from non-CSI. (Which is to say, I, as a software developer, create CSI all of the time, but I am of CSI.) I, therefore, believe that CSI defines the heart of the abiogenesis challenge.
The next question, however, is once CSI organisms exist, can they, through the mechanisms of RM+NS generate more CSI? We know that CSI (me) can generate more CSI if that CSI is capable of abstract reasoning.
I honestly suspect that the answer to this question is Yes, RM+NS can generate CSI. If so, this definition of CSI is useful in:
* Defining the abiogenesis challenge. How do we get from non-CSI to CSI?
* Providing a clear distinction betwee "life" and "non-life" - Life is CSI, though not all CSI (software) is life.
* Yet growth of CSI (as defined here) does not necessarily preclude "not of nature" because a mechanism RM+NS is at least suspected of being able to increase CSI.
IP: Logged
|
|
Irving
Member
Member # 535
|
posted 16. March 2006 18:12
quote:
I don't know where you get this. I am not seeking a definition that "does not rule out nature", I am only seeking a definition that fits the data rather than fitting a philosophy. If, by time you have a definition that fits the data, some philosophy is the only viable explanation, well, so be it.
Well, there's the problem again. The effort in defining CSI is NOT to fit the data. As I mentioned before, the effort is to define a characteristic that differentiates Design from Nature... and only then, see if Life exhibits that characteristic.
quote:
I fail to understand, then, why Dembski would suggest that Behe's "IC" concept is required to establish specified complexity.
I don't believe that Dembski has claimed that IC is required to establish specified complexity. He's claimed that IC is an "example" of specified complexity.
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 16. March 2006 18:25
Irving, you, and probably Dembski, have a different purpose, so are seeking a different (overloaded) definition than I have.
I am happy to let you and Dembski have a term which means "could not possibly happen by natural means", then use it to determine if anything in nature fits the definition.
I want a term which places DNA and computer software into the "in" group, but places the entire abiotic world in the "out" group. I want this because there seems to be an incredible confusion about what life is, an attempt to believe that some life is "simple".
I was of the understanding that Dembski's CSI was exactly that. I am dissappointed that it does not seem to be. I also have a lot more respect for those who say "I don't believe that Dembski's CSI exists."
I think I must create a new term to define what I want, just because the overloading is too confusing. So let me re-label my definition for clarity. Let me call it complex blueprint information or CBI:
quote:
CBI is any packet of information containing: information (order) with the nature of a blueprint (an independant detatchable pattern) and sufficent complexity (having limited compressability even knowing the nature of the data) to extend beyond UPB.
Now, can I own that term, and let you and Dembski own CSI?
IP: Logged
|
|
Irving
Member
Member # 535
|
posted 16. March 2006 18:28
quote:
The next question, however, is once CSI organisms exist, can they, through the mechanisms of RM+NS generate more CSI?
I would suggest that the answer is NO.
CSI is the upper limit to RM & NS. The problem with RM & NS, from a Popper standpoint, is that supposedly RM & NS can create anything. As a Theory of anything, is it really a Theory in a Popperesque sense? As a thought experiment, give and example of a single object that RM & NS can't create.
If there is such an object, then RM & NS isn't all powerful and isn't a Theory of Anything. CSI is an attempt to define a characteristic of just such an object. So, in a way, falsifying CSI is confirming RM & NS, and falsifying RM & NS confirms CSI...
If CSI is a valid concept, and RM & NS can't produce it, then both Theories move to a better scientific standing.
IP: Logged
|
|
|
Printer-friendly view of this topic