|
Author
|
Topic: Validating Design Discrimination Methodologies
|
RBH
Member
Member # 380
|
posted 17. October 2002 03:03
I'm going to fight to keep this thread on topic: "Validating Design Discrimination Methodologies"! ![[Smile]](smile.gif)
Mike Gene wrote quote:
Since the focus of MDT is on the designers, I'm not sure it will offer any real payoff when focusing on biotic reality. My emphasis is on using an ID approach to better understand the biological world . MDT's emphasis seems elsewhere, only paying lip service to the biological world as a springboard to focus on the designers.
Anyway, enough of these general points. What I will do instead is focus on another biological problem and again use my ID hypothesis as a research guide. I'll comment on MDT to see how well its ability to generate hypotheses compares to those my approach raises. For extra bonus, I'll even throw in the research potential that would stem from the blind watchmaker approach. Put simply, we need a concrete biological example with which to compare and contrast.
The several weeks since I proposed MDT have been interesting for me. I actually did start my thinking about MDT with a sole focus on the designers, since the proposed detection of intelligent design by Dembski's explanatory filter immediately raises the question of the designer(s). While the leading IDsts of that 'school' of ID (SDDID) have persistently denied that questions about the designer are relevant at this stage of development, I can't help but think that's simply avoidance. I continue to believe that question cannot be avoided in favor of the mere detection of design. As I've developed my thinking, and as others have commented, it has become clearer and clearer to me that the question of design and the question of designer(s) are inextricably woven together.
This thread, which is titled "Validating Design Discrimination Methodologies," reflects my growing realization that the methods by which one purports to detect design are tightly bound up with one's conception of designer. There is no 'neutral' design discrimination methodology. Any methodology is informed by theory, and that theory, if it is to provide historical narratives that explain any design that is detected, must include reference to 'designer(s).' Methodology flows from theory, and vice versa; they are inseparable.
The method by which one explores design in biology is governed by assumptions, implicit or explicit, about the designer(s). If one has one conception about designer(s), one uses one methodology. If one has a different conception, one uses a different methodology. If one thinks of white light as undifferented radiation, one does research on brightness and dimness and designs light meters to measure fluctuations in intensity and explores the functions of on/off switches. If one thinks about white light as a composite of colors, one uses a prism, and out of that flows Newton's Optics and ultimately the spectroscopes by means of which we study the composition of the stars.
That is why my attention has shifted in part to the question of the validation and reliability of designer discrimination methods. As I've written on a number of occasions now, if one takes MDT as a base, the methodological emphasis is very different from that implied by SDDID. The latter treats design, in my metaphor, as undifferentiated radiation with just one variable, intensity, which for SDDID has just two values, on or off. MDT treats design as a more complex composite phenomenon.
+++++++++
Now a few remarks on the development of MDT design discrimination methodology. I will first repeat what I said above: MDT's design discrimination methodology does not yet exist! As I have said before, and I will say again, an early focus of the MDT research program must be on methodology. Let me expand on how that program to develop a designer discrimination methodology might proceed.
We know that humans can discriminate among design themes in a number of domains - art, literature, software, and so on. It is absolutely correct that those discriminations depend on extensive background knowledge of the designers, of their design techniques, of the medium, and (to some extent at least) of the purposes and intentions of the designers. Human designers produce artifacts and other human beings make discriminations among those artifacts on the basis of their background knowledge. Thus those discriminations are not researcher-independent; they depend on the knowledge of the human judges.
Consider, though, one route I have previously sketchily proposed by way of methodology development. Suppose we require a group of programmers to write a large collection of C++ programs. Some are written by multiple programmers, the others by one or another single programmer. We assume that there was minimal top-down control in the writing of all of the programs. Only the broadest of controls - communication protocols among modules within the programs - are pre-defined for the programmer(s). Within those broad constraints, the programmers write their assigned program modules as they please.
As has been suggested before, a human judge versed in C++ could look at the source code of the programs and with pretty high reliability tell whether a given program was written by a single author or by multiple authors. It is absolutely clear that the human judge's discrimination utilizes background knowledge. The core question in developing a design discrimination methodology is whether one can create a 'mechanical' discriminator that can look at the same programs and, without being given access to any of the background knowledge the human judge has, find the properties of the programs that inform the human's discriminations. That is, can a mechanical discriminator itself "learn" to make the same discriminations the human judge did when the mechanical discriminator has only the array of programs to look at? Can it learn to recognize the characteristic 'signatures' of multiple designer programs as distinguished from single designer programs without the human judge's background knowledge, using only the properties of the artifacts? Is there enough information (in the technical and/or the ordinary language sense) in the artifacts to support reliable discriminations? That is what is required if design discrimination techniques are to be researcher-independent.
One avenue for the development of an MDT methodology I thought about was to attempt to adapt Dembski's algorithmic explanatory filter. But the EF fails to meet either the researcher-independence requirement or the reliability requirement. Obtaining a positive design inference from the EF depends on the ability of the human operator of the filter to eliminate all "regularity" explanations. Since we're not omniscient we can never know all the potential "regularity" explanations and thus a positive design inference produced by the EF is always subject to invalidation by "further notice" from those doing naturalistic research in biology. A positive design inference may be invalidated at any moment when an assistant professor of biology somewhere, desperate to make tenure, finds a "regularity" explanation for an Event that was previously inferred by the EF to be designed. That positive design inference instantly becomes a false positive. In fact, the moment it was made that design inference was a false positive; we just happened to find out about it later. How much later we find out about it depends on the work of naturalistic scientists; we may find out tomorrow, we may wait a hundred years. But the vulnerability is constant.
Even a merely conjectural naturalistic "regularity" explanation, if it is even only faintly plausible, must invalidate a positive design inference generated by the EF, since a plausible though untested naturalistic "regularity" explanation must have some non-negligible probability relative to the cosmically tiny UPB and therefore the design inference fails. (In slightly more technical terms, the assumption of a uniform PDF over all chance and regularity alternatives is invalidated by finding a plausible though untested naturalistic hypothesis because the appropriate PDF for estimating the probabilities for the EF algorithm then has a 'hump' - a mode - over a "regularity" alternative and the algorithm is sidetracked before it reaches the design inference.)
Thus the explanatory filter cannot mechanically detect design and it is unacceptably vulnerable to false positives. Therefore the algorithmic steps of the explanatory filter cannot provide a basis for the development of a multiple designer discriminator. Any extension of the filter for that discriminator will suffer from the same disqualifying arguments.
MDT proposes to attempt the development of automated, mechanical designer discrimination methods that are based solely on the properties of artifacts themselves. We know methods - automated mechanical methods (relatively) untainted by domain-specific human background knowledge - that enable machines (computers) to learn to make subtle and difficult discriminations in real-world environments. They did not learn those discrimination skills from humans - they weren't programmed to make the discriminations. In fact, the discriminations that the programs make are impossible for all but a few humans in the world, even though the environmental stimuli that the programs.attend to are as detectable by humans as by the programs. The programs 'look' at phenomena - descriptions of the behaviour of complex adaptive systems are piped into the computers from the outside world - and make discriminations that most humans cannot. Those programs are exist, they are running at this moment in real machines, and based on their discriminations they autonomously make decisions that control processes worth tens of millions of dollars daily. In the rapidly advancing machine intelligence world we may well find the pattern recognition and discrimination technology we need to mechanize the multiple designer discrimination problem.
RBH
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 17. October 2002 04:58
Jack writes to me,
quote:
I find it strange that persons that can't tell us what they would consider evidence for ID are discussing methods for detecting it. If you don't know what evidence for ID would look like then it could be right in front of you and you wouldn't recognize it.
Jack - I have told you, and have in fact written extensively about, the type of evidence I would like to see about design - evidence that would pass through Dembski's filter by showing that the probability of something having happened through purely natural causes is improbably low.
You don't seem to accept that answer, but it is an answer based on ID theory.
Secondly I have told you that merely suspecting design is not a strong enough guide for me - "suspecting design" has been full of "false positives" over the centuries. I would like to see more solid evidence than that.
I am not looking for all the excessively strong things you mentioned in your earlier posts - interviews with the designers, etc. I just would like to see us develop some methods that will produce some evidence of the type accepted in science - evidence based on measurable quantities gathered through reproducible means. I have suggested ideas for developing such methods.
What more do you want?
[ 17. October 2002, 11:06: Message edited by: Evan ]
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 17. October 2002 05:17
In response to my comment about why he thinks MDT is religious, Jack Foster writes,
quote:
RBH has said that the multiple designers must be unembodied. But the only evidence that we have for the designer is the design; so how do we distinguish between the embodiment or the unembodiment of our designers?
Faith. That's the only way that I can see it. And faith takes us out of the realm of science and into the realm of religion.
First, I don’t believe RBH said that the designers must be disembodied. What he said was that his theory was going to follow Dembski and assume they were disembodied. Both he and I have pointed out that we understand that design theories where the designers are embodied, such as ET theories, offer different hypotheses about the source of design, but that’s not the variation we are interested in or writing about.
Secondly, you seem to be saying that any design theory which posits disembodied designers, either as an assumption or through inference based on evidence, is religious. This is probably a subject for a different thread, but that is a strong statement. Dembski clearly states that he assumes the designer is disembodied, and even offers ideas about how that disembodied being interacts with the physical world. Does that mean his theories are religious?
More empirically, Jack Foster asks “how do we distinguish between the embodiment or the unembodiment of our designers?”
Well, here’s an example from the ARN thread that mentioned cichlid fishes in an African lake. Let us suppose that empirically based studies showed that at least some aspects of the diversification of those 400 species were designed (at least some design theorists hypothesize that natural causes are insufficient to produce speciation.)
Given what we know about what type of embodied creatures have been on earth in the past 13,000 years, which would be the more reasonable hypothesis: that disembodied designers were responsible for that design, or embodied ones? It certainly seems to me that the evidence would point to something other than embodied designers.
In this case inferring the existence of disembodied designers would be reasonable, and there would be nothing religious about it.
[ 17. October 2002, 11:09: Message edited by: Evan ]
IP: Logged
|
|
Jack Foster
Member
Member # 79
|
posted 17. October 2002 11:42
Hi Evan:
quote:
First, I don’t believe RBH said that the designers must be disembodied. What he said was that his theory was going to follow Dembski and assume they were disembodied.
MDT does NOT follow Bill Dembski. Bill is a Christian, and believes by faith in a single designer. Nowhere is this faith a part of his scientific position. In contrast, RBH lists unembodiment as the first property of the Multiple Designers in MDT:
II. Some Properties of Multiple Designers
A. The multiple designers are unembodied.
Now RBH has begun to place MDT in contrast with something he calls "SDDID" for Single Designer Dembski ID. This is not a construct of Bill Dembski; it's a red herring invented by RBH. There's no such thing as SDDID.
quote:
Given what we know about what type of embodied creatures have been on earth in the past 13,000 years, which would be the more reasonable hypothesis: that disembodied designers were responsible for that design, or embodied ones?
We have no way of knowing, or more importantly . . . of testing. We currently have no concrete evidence for the designer, except for the design. You can speculate about the nature of the designer if you care to. But until you have some way of testing the speculation; it ain't science.
By contrast, Bill Dembski's design-centric ID (NOT SDDID) is within the realm of Science. If ID were to succeed in falsifying naturalistic abiogenesis, for instance, then chance's logical complement - design - would be scientifically implicated. Success in the task would change the direction of Science.
[ 17. October 2002, 11:44: Message edited by: Jack Foster ]
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 17. October 2002 12:11
From Dembski’s essay “ID Coming Clean” (which is a significant title):
quote:
The reason that attributing specified complexity to intelligence for biological systems is regarded as problematic is because such an intelligence would in all likelihood have to be unembodied (though strictly speaking this is not required of intelligent design--the designer could in principle be an embodied intelligence, as with the panspermia theories).
Furthermore, immediately following this quote, Dembski spends quite a bit of time discussing how a supernatural, unembodied intelligence could interact with the physical world.
This is all irrespective of whether Dembski himself religiously believes that the designer is the Christian God. Positing an intelligence which introduces design into the physical world necessarily brings up the question of whether the designers are embodied or not. I have never claimed that evidence for whether the designers are any particular God - that is religious.
But evidence that the designers are embodied or not is in theory ascertainable, and is not religious. Historical sciences make inferences based on the evidence all the time. If, (and I understand this is hypothetical, as the tools to determine this have not been developed - that is step 1)the cichlid fishes proved to be designed, as well as many comparable episodes of speciation that we know of, the lack of any evidence that embodied creatures have visited earth and done this work is certainly eliminative evidence that the designers are unembodied and work in ways that we don’t understand (along the lines of Dembski’s musings on this subject.)
So even though I, and RBH and Dembski, acknowledge the possibility of panspermia or ET theories of design, the hypothesis that the designers are unembodied and work directly on life on earth is quite within the realm of ID theory, and is not a religious addition to the theory.
IP: Logged
|
|
Jack Foster
Member
Member # 79
|
posted 17. October 2002 13:11
I'll grant you that speculation regarding the nature of the designer does not have to be religious. But of course that's true for any religion as well. The ideas of any religion can be repositioned as scientific speculation, and in fact, that's what Scientific Creationism is: a repositioning of Christianity as scientific speculation.
On a different topic, I think the use of the acronym and the term SDDID -- Single-Designer Dembskian ID is inappropriate, . . . especiallly when placed in direct contrast to MDT. It falsely attributes a quality to Dembski's work. His ID is (appropriately) design-centric, and the nature (and quantity!) of the designer is not part of his scientific writings. I would ask RBH and others to refrain from using the term, as it is misleading; or alternatively to get permission from Bill Dembski to use it.
The more I think about this, the more it bothers me. It borders on dishonest, and is easily remedied.
RBH, please revert to your use of "SDT -- Single Designer Theory". (Though still . . . nobody is proposing SDT. But at least there's no false attribution.)
[ 17. October 2002, 13:16: Message edited by: Jack Foster ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 17. October 2002 14:05
Hi, Jack F. You wrote quote:
On a different topic, I think the use of the acronym and the term SDDID -- Single-Designer Dembskian ID is inappropriate, . . . especiallly when placed in direct contrast to MDT. It falsely attributes a quality to Dembski's work. His ID is (appropriately) design-centric, and the nature (and quantity!) of the designer is not part of his scientific writings. I would ask RBH and others to refrain from using the term, as it is misleading; or alternatively to get permission from Bill Dembski to use it.
How then can I refer to it in a way that distinguishes it from the wide and varied array of views that are included in the ID "big tent." I use "SDDID" for the following reasons:
"SD" means "Single Designer." I explained quite clearly in the OP of the Introduction to MDT thread what the motivation was for believing that a single designer is at the heart of the Discovery Institute's Senior Fellows' (note the plural possessive) view of ID.
The second "D" stands for "Dembskian," meaning that insofar as it appears relevant and useful, MDT uses Dembski's texts, especially TDI, NFL, and the "Coming clean" essay, as sources for what I have and will continue to contrast MDT with. It has been argued directly to me (in personal communications from a spokesman for an organization of ID proponents) that teaching ID in Ohio high school science classes is a good idea because it forms a foil for teaching evolutionary theory, thereby strengthening the teaching of evolutionary theory. Similarly, Dembski's published views of ID provide a foil, a comparative reference, by means of which to clarify and develop MDT both theoretically and methodologically. Dembski's works contain the clearest methodological descriptions in design theory. That view is also among the the most popularized versions of ID around and hence is the most visible intellectual competitor to MDT.
"ID" takes the meaning generally attributed to it: "intelligent design," where "intelligent" means (again as Dembski says) making choices. There is also a strong thread of thought to the effect that the choices take into account the designer(s) being intentional agent(s), and "design" means "not produced by purely naturalistic causal variables but by the action of an intelligent agency or agencies acting on intentions." Hence, SDDID.
I suppose I could amend the acronym, using, e.g., SDDIID, for "Single Designer Discovery Institute Intelligent Design." I'm indifferent to the acronym used so long as it clearly indicates one fairly clear stream of ID thought among the plethora of views. But I have learned from the critical postings here and on ARN that just "ID" or even "SDT" by itself is unacceptably vague and non-specific.
Jack Foster wrote in an earlier posting: quote:
MDT does NOT follow Bill Dembski. Bill is a Christian, and believes by faith in a single designer. Nowhere is this faith a part of his scientific position. In contrast, RBH lists unembodiment as the first property of the Multiple Designers in MDT:
II. Some Properties of Multiple Designers
A. The multiple designers are unembodied.
Now RBH has begun to place MDT in contrast with something he calls "SDDID" for Single Designer Dembski ID. This is not a construct of Bill Dembski; it's a red herring invented by RBH. There's no such thing as SDDID..
SDDID (or SDDIID, if you prefer) is emphatically not a "red herring. As I noted above, it is an attempt to specify a fairly specific version of ID out of the vague and amorphous congeries of speculations, notions, and conjectures that inhabit the ID "big tent."
As Evan has clearly asserted, there are no religious motivations for the proposal and development of MDT, and the assumption of "unembodied" designers is based directly on Dembski's version of ID. Similarly, that Dembski is a Christian is not, as he himself at least strongly implies, a relevant factor in his development the Explanatory Filter, Complex Specified Information, or any other intellectual apparatus he describes for use in ID as a scientific enterprise. Nor does Behe, to take another example of a DI Senior Fellow, depend on his Roman Catholicism to define irreducible complexity; he depends on his knowledge and interpretations of biochemistry and evolutionary theory. While criticizing the definition, utility, or validity of ICness as an indicator of design on logical, empirical, or theoretical interpretation grounds is perfectly legitimate here, in this forum criticizing it on religious grounds is not.
Raising religious objections to, or potential religious implications of MDT in the context of this board and these discussions is the real red herring.
Jack Foster further wrote quote:
We have no way of knowing, or more importantly . . . of testing. We currently have no concrete evidence for the designer, except for the design. You can speculate about the nature of the designer if you care to. But until you have some way of testing the speculation; it ain't science.
We currently have no evidence for either the designer(s) or the design(s). We have a scattering of anecdotes, but I do not know of one single published example of the systematic application of the Explanatory Filter to a biological phenomenon where the EF is formally applied in the way TDI and NFL assert it should be formally applied. There are informal 'demonstrations' and hypothetical examples, but no systematic research. There are no comparative data showing classes of phenomena producing inferences of design versus inferences of chance or regularity. The several test beds appropriate for developing and testing design discrimination methodologies that have been identified - human designs of various sorts, interesting physical systems like the Oklo reactor, and biological phenomena like the Lake Victoria cichlids - all lie fallow, waiting for an enterprising ID researcher to use them to validate and calibrate the methodology - the intellectual instruments - of design theories.
I have attempted to first sketch and then begin to flesh out the way an MDT designer discrimination technology might be developed in order to seek the kind of evidence Jack Foster asserts we do not have. I have suggested some current human designer discrimination technologies that might be adapted to that task, and there are others that have been suggested to me recently, as for example adapting the Zip compression method for author attribution and speech recognition. I am not as familiar with that work as I am with other approaches to machine pattern recognition and discrimination so I won't discuss it further. MDT is at least looking carefully at methodological issues and is attempting to map routes to the development, validation, and calibration of methodologies that could yield real evidence.
RBH
Edited for some typos. Any that remain are the responsibility of the IP transmission protocol!. ![[Wink]](wink.gif)
[ 17. October 2002, 14:11: Message edited by: RBH ]
IP: Logged
|
|
Jack
Member
Member # 265
|
posted 17. October 2002 17:03
Hi Evan,
I posted my reply to you on the "ID research program" thread since it's more on topic over there.
IP: Logged
|
|
Jack Foster
Member
Member # 79
|
posted 17. October 2002 18:08
The problem, Mr. Moderator and RBH, is that the label conflates Bill Dembski's monotheistic religious views with his design-centric ID views. Perhaps someday Bill will want to pursue a designer-centric hypothesis. But he isn't now, and RBH's label implies that he is.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 17. October 2002 21:09
Jack Foster wrote quote:
The problem, Mr. Moderator and RBH, is that the label conflates Bill Dembski's monotheistic religious views with his design-centric ID views. Perhaps someday Bill will want to pursue a designer-centric hypothesis. But he isn't now, and RBH's label implies that he is.
I'm sure Dembski can speak for himself in this respect. I'm aware that he has explicitly asserted that he thinks conjectures, hypotheses, or theories about the designer(s) are premature, and that design detection should be the primary focus. Further, I am aware that a designer-centric hypothesis is not currently an explicit component of his approach to ID.
However, as I noted in the first paragraph of the OP of the Introduction to MDT thread, there is a strong default to, and assumption of, a single designer in Dembski's writing. A very brief analysis of "Intelligent Design Coming Clean" will serve to illustrate. The word "designer" occurs in the singular 51 times in that essay, while the plural "designers" appears 12 times. Of the 12 times the plural "designers" appears, 7 are in the context of three paragraphs of critique of Elliot Sober, and appear to refer more or less directly to human designers, though a couple of those references are ambiguous. I find no clear instance of the singular "designer" referring to a human.
Similarly, "unembodied intelligence" appears four times, all in the singular. "Intelligent agents" appears four times in the plural, all four referring specifically to humans, and once in the singular ("intelligent agent") in a context which does not refer to a human.
Further, Dembski writes "Once it is settled that certain biological systems are designed, the door is open to a new set of research problems. Here are some of the key problems:" The last five on the list are quote:
Separation of Causes Problem -- How does one tease apart the effects of intelligent causes from natural causes, both of which could have affected the object in question?
For instance, a rusted old Cadillac exhibits the effects of both design and weathering?
Ethical Problem -- Is the design morally right?
Aesthetics Problem -- Is the design beautiful?
Intentionality Problem -- What was the intention of the designer in producing a given designed object?
Identity Problem -- Who is the designer?
(all italics original)
Dembski goes on immediately to say quote:
To be sure, the last four questions are not questions of science, but they arise very quickly once design is back on the table for serious discussion. As for the other questions, they are strictly scientific (indeed, many special sciences, like archeology or SETI, already raise them).
Note particularly the singular "designer" in the last two. Although the list is labelled "a new set of research problems," Dembski classifies the last four as "not questions of science." But one man's fence is another man's bridge.
My main point is that there is a powerful default to the singular designer assumption (added in edit: and that default is not methodologically neutral). MDT is a wholly legitimate extension of the design discrimination problem itself, whether it is a discrimination between designed and not designed or discriminations among designs, and MDT is a logical alternative to the default single designer assumption in mainstream ID. And yes, questions of identity(ies) and intentions do arise very quickly once design is back on the table, and the empirical study of designed structures, processes, and objects is an avenue to addressing those questions, particularly the identity(ies) question. As with the "Separation of causes" question, that can be studied scientifically given the appropriate methodology for discriminating among design(s).
If it will help Jack Foster feel better, I'm willing to call it SDDIID. That distinguishes it from the other streams of ID thought sufficiently to make it clear (I hope!) which version of ID is being referred to.
RBH
edited for typos
[ 17. October 2002, 21:14: Message edited by: RBH ]
IP: Logged
|
|
Moderator
Administrator
Member # 1
|
posted 18. October 2002 00:18
RBH,
I would like for you to give me a concise three sentence explanation of the difference between SDT and SDDID.
I will then rule on whether I feel that your use of the more descriptive acronyms is antagonistic or not.
[Edit: I retract the following]
At the moment, my sense is that you are sneaking in a pot-shot. But I'll give you a chance.
[Edit:]
Though I would advise RBH to be more general and descriptive in his use of acronyms in the future. Why choose SDDID versus SDUID?
SDUID seems to be far more descriptive than SDDID.
[ 18. October 2002, 08:55: Message edited by: Moderator ]
IP: Logged
|
|
Jack Foster
Member
Member # 79
|
posted 18. October 2002 03:51
One last, short comment from me. I don't think RBH is sneaking in a pot shot. What's going on is that he's defining Dembskian ID as S(U)DDID . . . the U for unembodied. This is clear from his description of MDT, which he claims to model after S(U)DDID. Perhaps the better acronym would be GOD-ID. No pot-shot there.
The question is this: Is Bill Dembski willing to endorse this definition of his version of ID? I'm actually quite interested in the answer.
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 18. October 2002 10:48
I am somewhat baffled by this acronym discussion. All sorts of acronyms proliferate, and people make them up all the time to help facilitate discussion.
From the beginning here at ISCID , I have often needed to make it clear that my interest in ID is from Dembski’s viewpoint (or at least my understanding of it.) There are many different, and fairly vague, versions of ID. One of my interests has been in making the theory more specific, and in order to do so I have chosen to focus on one version - Dembski’s - as opposed to others.
It seems to me clear that Dembski’s version involves a singular designer, unembodied, who is outside of nature (transcendent) and acts on nature (as opposed to an immanent intelligence such as might be posited by the neo-Gnostics or the Endogenous Adaptive Mutationists.) This is is the version of ID I am interested in (except that I have accepted the argument put forth by RBH that multiple designers should be considered.)
Therefore Multiple Unembodied Designer ID that incorporates the basic ideas of Dembski (EF, CSI, IC, and NFL - how’s that for acronyms!) is the version of ID being discussed.
So I don’t see any problem at all with MUDDID - Multiple Unembodied Designers Dembskian ID, or some other combination of these letters.
[ 18. October 2002, 11:29: Message edited by: Evan ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 18. October 2002 11:03
The Moderator wrote quote:
I would like for you to give me a concise three sentence explanation of the difference between SDT and SDDID.
This is as concise as I can make it and still be at least minimally fair to the complexity of the issues.
Three sentences:
Single-Designer Theory (SDT) assumes a single intelligent designing agency, with no other defining ID theoretical or methodological qualifications or assumptions specified.
SDDID (or SDUID or SDDIID) assumes a single intelligent designing agency to which we do not have direct observational access, which is unembodied, intermittently interventionist through time, and its designs are discriminable from the non-designed via the methods described in TDI and NFL (which differ substantially from the methodologies of other ID research programs like baraminology); all of which have significant implications for design detection and discrimination methodologies and for the generation of theory-based hypotheses.
SDDID (or SDDIID or SDUID) accepts (to some variable extent depending on the particular theorist) common descent, explicitly or implicitly excludes some major classes of ID hypotheses (such as biologically pervasive inherent intelligence, and front-loading), and appears to be supported (as far as one can tell) by several of the principal scientific and theoretical exponents of intelligent design.
Some elaboration:
Clearly, the major point of unease and contention is the assumption that the intelligent agency is unembodied. That assumption has critical implications for methodologies and research program directions. A major implication is that it forces research attention to analyses of only the (biological) artifacts hypothesized to have been produced by the designer(s). It excludes both a research program designed to search for other kinds of signs and evidence that might have been left by the intermittent physical presence of embodied intelligent designing agent(s), and a research program that seeks direct observational access to the embodied designer(s).
I can live with "SDUID" or "SDDIID". As I said, I am not wedded to a particular acronym. What I want is to clearly distinguish the particular version of ID I mean, very briefly summarized above, from the confusingly wide array of different versions of ID out there in the world, in order to have a reference point for MDT. As I have also said, I see MDT as a logical theoretical and empirical extension of (SDDID or SDDIID or SDUID) as that position has been presented in a number of major publications in the last 5 years.
RBH
Added in edit: The Moderator also asked why I used "D" rather than "U". I didn't answer answer that question above.
I originally used "D" rather than "U" for two reasons. First, the thought of a choice didn't even cross my mind because I didn't see the assumption of "U" as a terribly contentious issue in anything I have read in the ID literature. Second, in a number of publications Dembski routinely refers to an unembodied designer, answers objections by providing hypothetical means by which an unembodied designer might transmit CSI to material form, and explicitly discusses the relation between an unembodied designer's activities and natural laws. He is the major theoretician and clearly the most widely published spokesman of ID to the 'outside' world. It seemed an uncontroversial choice to me. But, as I've repeated several times now, I'm not wedded to it.
[ 18. October 2002, 12:40: Message edited by: RBH ]
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 18. October 2002 11:58
Very good, RBH. I think that it is useful to make explicit some of the ways in which an assumption of unembodiment will effect a research program, as opposed to an assumption of embodiment.
If one is hypothesizing embodied designers, one will look for evidence of their bodily presence on earth apart from the biological effect they have had.
If one is hypothesizing unembodied designers (who interact in undetectable ways, as described in ID Coming Clean and elsewhere) one will look for anomalous (that is, improbable) events right at the genetic level (or perhaps elsewhere).
Definitely, research methodologies will vary significantly depending on whether one is looking for evidence of embodied or disembodied designers.
I also appreciate RBH’s concise summary of other aspects of the theory. Let me make more explicit two other issues.
1) Dembski makes it clear, and I fully accept this assumption, that the acts of design are not miraculous - the laws of nature are not suddenly and discontinuously contravened when design happens. (See ID Coming Clean, for example). Rather, information (CSI) is imparted into the world in an undetectable way - undetectable in the sense that no one act itself breaks any natural law. It is the cumulative effect of many (how many?) such acts that adds up to the improbability that distinguishes design. That is why calculating the probability of an event that has a history is the key to detecting design.
In his preface to NFL (I don’t have the quote right here) Dembski says that the fundamental property of the designer (singular) is that it is able to, and does, impart CSI into the world. That is the fundamental assumption about the designer.
2) The other issue, which Dembski does not clearly address, I think, is where does the designer (or, where do the designers) interact with the world. Do they have the power to influence any type of event, or just biological events. Can they influence large scale events (such as the success or failure of an entire population of organisms), or are they limited to, perhaps, quantum events where the manipulation of probabilities can possibly take place in an undetectable way.
Or could they influence large scale events by manipulating a huge number of quantum events simultaneously?
Again, how one hypothesizes about this will effect the development of a research program.
This topic is beyond the scope of this thread, and I don’t mean for it to take us off topic. But since RBH did such a good job of summarizing some of the issues that I thought it might be useful to mention these other issues.
As far as acronyms go, I’m for keeping it simple. My position is that we are discussing Dembski’s ideas about ID, with the additional hypothesis that the designers might be multiple. The fact that these designers are assumed to be unembodied does not really need to be a part of the name, because that is subsumed under the adjective Dembskian, I think.
[ 18. October 2002, 12:19: Message edited by: Evan ]
IP: Logged
|
|
|
Printer-friendly view of this topic