|
Author
|
Topic: Does intelligence imply “motive”?
|
gedanken
Member
Member # 594
|
posted 29. August 2003 11:06
Does intelligence imply “motive”?
I’m concerned that an important aspect of recognizing “intelligence” is being left out of ID analysis. The general category I am concerned with is that “intelligence” per se is clearly a characteristic being used to classify an otherwise unknown or unspecified mechanism for an observed result.
ID concepts have analyzed what is considered to be aspects that indicate “intelligence”, but do not usually go further to deal in details of mechanism. If ID is to be a truly “scientific” concept, it must in my opinion move to more details of mechanism. If it does not, it can’t possibly expect to be accepted as scientific “theory”, having in fact no theory of how the events come about but being rather a classification scheme only for events.
What I am asking is to delve more deeply into the characteristics of intelligence, and how it effects physical events in the real world.
In “Caputo” example (commonly used to illustrate ID concepts), we have a clear side-indication of “motive”. Caputo clearly had a motive to cause the sequence observed of almost always choosing one party for the top of the ballot, while claiming to use a 50:50 “random” selection method. Yet the explanations of the Caputo example ignore this, and leave the issue of “motive” as a background aspect that all the readers understand but not noting that it is actually part of the ID analysis itself. But in this case we have a clear case of mechanism being understood in the analysis, though not presented explicitly. Such presentations may falsely imply that the ID classification method is working, because all the examples have this underlying characteristic that is understood by the reader, yet not listed as part of the technique itself.
Without that element of “motive”, we could have considered that the cause might have been more likely as a mechanical failure. Consider a stuck device that was supposed to select at 50:50 ratio, but caused unintentionally. It is the aspect of “motive” that makes the Caputo case seem to be a clear case of ID, not the fact that an independent string of D’s would be “specified”. In fact I think that the motive is a clear, yet unstated aspect of the claim of “specification” that cannot be ignored. Other cases a sequence of repetition of the same symbol would be considered extremely uninteresting as opposed to be a “specification” – I think the issue of “motive” (unstated) is highly involved in making the series of Democrat selections have the character that ID concepts try to capture as “specification”.
I am requesting that such conceptual work should begin to include aspects of mechanism in the ID event, so as to give some greater degree of scientific content to ID analysis.
Now I’m not saying that “motive” has been left out of all ID presentations. For sure Mike Gene has given analysis that included “motive”, as have others. My point is that “motive” is one in a long chain of “mechanism” issues that should be included.
The “motive” aspect of mechanism is most certainly not sufficient by itself to make ID concepts appear more scientific to the general science community, I’m just saying that it is a step. In the long term, the improvements needed to make ID understandable as “science”, I would suggest that many other mechanisms of intelligence need to be considered in the analysis.
[ 29. August 2003, 11:11: Message edited by: gedanken ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 30. August 2003 06:05
While I am not certain that intelligence always implies a discernable motive, I do agree that finding motive is of great help in our everyday task of ascribing events to intelligent or nonintelligent causes.
Specifically, finding motive helps us skip the nearly impossible step of considering all law- and chance-based processes, by finding a single feature (motive) that correlates highly enough with intelligent action that the chance that the event is not intelliently caused can be safely neglected (in the absence of some other strong indication against intelligent cause).
This is a particularly important step, it seems to me, because there are so many merely complex yet structured patterns that we would wish to call unintentional (such as a spectrogram of wind noise in a canyon), and because our descriptive powers are so extensive that, with enough study, it seems plausible that we could construct something that looks like a specification for nearly anything that isn't a uniformly distributed random process.
It isn't clear to me, however, how to apply motive when the designers aren't around to explain their motives some of the time. Humans are good at anthropomorphizing observations: I have been known to say things like "the atom wants to return to the ground state", while not intending to imply that atoms have motive. Perhaps we can avoid a similar error when it comes to implying that microorganisms intend things ("the bacterium intends to move, and wants a flagellum to do so"). But for any postulated superdesigner, it's difficult for me to see how we could do better than to trade inventing specifications with inventing motives.
If I were to guess, I'd guess that we were even better at the latter.
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 30. August 2003 08:19
Rex, I’m quite sure that the concept of motive could also be highly abused. So there is no assurance that any improvement in making ID claims seem acceptable to science might occur. That is a matter of “culture”, in that attempts to recognize the actual underpinnings of nature studied in science should not be forced into a pattern that the viewer wants to find as opposed to what is actually there. If a researcher “invented” motive that would be no more “scientific” than if the researcher “invented” an apparent specification claim, wherein either was poorly defined and thus subjectively ascribed to make the apparent situation fit the desired result.
The reason I bring it up is that I was examining examples used in explaining why ID is thought or claimed to work. I have yet to find any of these examples in which “motive” was not obviously considered, even though it is explicitly excluded by the formally described method of ID.
So if ID pattern classification method is claimed to work by a certain formal sequence of steps, yet practitioners in actually use a different sequence of steps that is different from the formal sequence, then we would easily have a case of bias in demonstrating supposed correctness of the method. This would be a classic case of false claim of some procedure working, as we know that science is filled with examples in which the researcher wanted a particular result to be true and biased his method to cause the result. That is why peer reviewed mainstream literature is so important, because someone will readily notice such results and point them out, confirming by constructing other demonstrations without that bias that demonstrate the falsehood of the original claim. It is also why blind and “double blind” methods are very useful in many scientific tests, as they reduce the false correlation effects of subjective elements.
So my point here may not be particularly useful in making ID recognition procedure more “blind” or non-subjective.
My main point, rather, is that it is easily and readily demonstrable that the “motive” was considered as part of the claim of “specification”, in terms of the examples given. They virtually all give the reader a background (non-stated) understanding through motive, yet the formal procedure does not mention the motive.
So this gives us an opportunity to observe how the actual use fails to be objective, precisely because we can so readily see how “motive” was incorporated in the examples used to supposedly demonstrate correctness of ID.
It also gives an opportunity to improve ID’s formal procedures in two aspects:
1) If procedure has motive incorporated, it may more accurately represent the ID procedure that is used in practice.
2) If examples are found to clearly demonstrate correctness of ID methods (not involving “motive”), then we might have better demonstration of correctness of such existing formal ID methodology.
Now I am not particularly confident that “2” will pan out in practice. I don’t know if such cases can be found that could show a correct working of the formal ID procedure, yet in which the so-called “specification” was not actually subjectively recognized by way of “motive”. This might actually appear as a downfall of ID. I don’t know the answer to this. But I have no particular bias that ID method should succeed to prove itself useful, my main interest is in understanding whether it can be “scientific”.
--
Added: Part of my interest here is that the ID claim is usually that the characteristics of the designer are not considered and don’t need to be considered. But if that is true, then case 2 should be applicable, that there should be ready examples of ID recognition demonstrably working, but without the subjective (and unspoken) consideration of “motive” being incorporated in the recognition procedure. I would like to see these examples given, or to see ID procedures updated to reflect how they are actually applied.
[ 30. August 2003, 09:48: Message edited by: gedanken ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 31. August 2003 01:56
For what it's worth, the SETI examples--at least inasmuch as they reference the real SETI efforts--don't seem to have much appeal to motive. Complex intelligent beings create complex output, not out of any particular motive, but as a side effect of being motivated to do anything at all. This needn't enter into our considerations when searching for the complex patterns created by complex beings.
Unfortunately, I don't think that this is worth very much, precisely because if the SETI effort did find a complex pattern, my immediate objection would be that we didn't actually know that it was created by a being rather than some physical process (e.g. the spectrum from the sun is quite complex, but we typically view that as a consequence of physical law, not intelligence). The easy way to resolve this objection would be to point out an immediate motive behind the message, e.g. to communicate with other intelligent life forms, which explains the structure of the signal. However, I think there are relatively motive-free approaches as well. For example, if we intercept a dialog and can make out linguistic structures, the parsimonius explanation is that it is in fact linguistic rather than some unintelligent process producing output with the statistics of language.
(Although advanced aliens ought to be well aware of information theory, and hence their communications would presumably be over maximally utilized channels, and therefore would be indistingusihable from noise. Oops. Anyway.)
In any case, I do think that there are ways out of the motivation trap--and I do consider it a trap, because I am far more concerned about bias in a method than describing mechanisms in a scientific way. It is difficult to use bias-prone methods for any rigorous work, and I suppose I am a stickler for rigorous work.
As my SETI example above illustrates, the other ways out (that I find compelling) also refer to human behavior--using language, for instance. I cannot think of any that would be of use in, say, answering questions about the origin of life. Indeed, I think there's ample evidence that there are some extremely non-human aspects of the "design" in living organisms: life is robust, self-repairing, self-replicating, and has no apparent external purpose. Things we design tend to be specialized, non-repairing, non-replicating, and for a purpose external to the object itself.
Intelligence involves being able to predict outcomes, yet we know now that predicting the outcome of setting transcription/translation machinery to work on a strand of DNA is horrendously complex. Given this, and the features of life described above, I wonder whether there is anything left in the mechanism of ID events that we can be sure of (in analogy with humans) aside from motive.
So perhaps there are not many other mechanisms of intelligence that we need to consider, and indeed, the one that we could be sure of may be too fraught with difficulty to actually use. I don't say this to discourage people from working in this vein, however! Quiet the contrary--I think the issue very much deserves further study. It is advisable to be aware of problem areas in advance, however.
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 31. August 2003 17:48
(Once again I point out that the addition of “motive” may not make for more productive ID analysis. Rather I think it is one of many details about the “designer” that are smuggled into the discussion of all examples that are used to claim that ID recognition procedures are “reliable”. I only bring it up because I think that the smuggling of “motive” aspect into the recognition of claimed “specification” is so readily demonstrable. To wit …)
The SETI is a good example in my opinion, but let’s examine it in more detail.
One of the most commonly given examples relating to SETI, given in popular explanations of ID, is the “Contact” movie example. The sequence there had complex pattern that was evidently designed for easy recognition by a very remote (in distance and communicating language) society. The sequence was a sequence of prime numbers, apparently designed for recognition as an intentional communication pattern. The clear “motive” in such a pattern is the ability of a receiving end to recognize the communication.
(All patterns that would have a synchronization or identification marker at their beginning would have “motive” of recognition by receiving party. We regularly send internet packets with a recognition header intended for the receiver to recognize the packet. Similarly for that communication starting with prime numbers – just a different level of what the receiving system is. Motive is the same.)
Then the aspects of what might be recognized by SETI: Apparent communications of any sort. Now this would be beyond simply recognizing a signal and claiming it was of intelligent design, rather what SETI might recognize would be a signal that has characteristics of communication. Here the recognized pattern is one of intended communication, the motive is to communicate, and possible patterns from that motive is what drives the specific methods of distinguishing signals of that sort from background “natural” non-intelligent cause signals. So in essence SETI is making very strong assumptions about the sources it is looking for, assumptions far beyond a generic intelligence per se. Only by making such further assumptions can SETI easily categorize the possible signals it is looking for.
Now of course Rex brings up that SETI would first be searching for a “complex” pattern. Looking for “complex” pattern is not looking for intelligence. So at that level, SETI is not an example of a so-called “scientific” investigation looking for intelligence. (Ignoring the name of the project, as this is a sub-level of the project being described.) Only after the apparently “complex” signal pattern has been first recognized as unusually complex by some measure would it be considered by the humans for whether it was an indication of some form of “intelligent” causation. It is in this level that I claim the “motive” aspect would be applied, as even such simple “motive” as signal being part of something communicated is sufficient to imply such motive aspect used in the analysis.
Additionally I don’t think that SETI is a competition to the methods of ID usually considered at ISCID, such as the “explanatory filter”. I don’t, for example, consider that SETI uses any version of the EF, certainly no version that ignores all details of the “designing” intelligence and considers patterns for intelligence in the abstract. People may argue with that, but I will claim that “motive” is one of many aspects that will be further considered – and we would have to present details to analyze the case in question.
Furthermore I don’t know of any successful examples from SETI of an actual confirmed recognition. So SETI could not possibly provide inductive examples to show correctness of application of design inference methods.
SETI makes myriad assumptions about the intelligence and how it might launch signals that could be recognized. It is not like the “explanatory filter,” claimed to be devoid of such assumptions, so it does not much affect the subject I am examining. In fact SETI is a very good example that inferences about intelligence that could be considered “scientific” indeed do consider details about the intelligent agent and mechanism for that agent to affect the event in question. It is this denial that such details of mechanism need to be considered that in part concerns me.
So to reiterate, if all examples used to demonstrate the “explanatory filter” or other design inference methods from the ID community smuggle the aspect of motive into the recognition process, while claiming that aspects of the designer are not being considered, then we have a clear demonstration that such design inference methods are not being accurately described. The people who describe these methods should help us by improving those descriptions, or by providing examples of those specific design inference methods.
After such improvement, or such widening of the examples of demonstrated correctness of applications of EF that don’t smuggle in motive, then we can examine them further to see if they have achieved any more “scientific” methodology.
So I agree pretty much with everything Rex said, even about what is of greatest interest. I only bring this subject up because of its obvious case of “smuggling” in motive in examples proffered.
[ 31. August 2003, 19:02: Message edited by: gedanken ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 31. August 2003 18:32
I thought I would add to this a partial example.
Now many hypothetical examples are used to illustrate design inference, not just actual physical examples. I am going to offer a hypothetical example that is not a case that is claimed to have design that is known, but rather one in which the cause is not known. We can consider whether the explanatory filter is helping us in this case.
A slot machine is designed so that it presents a series of 10 symbols that will be either a “R” or a “D”. Now the payoff occurs according to the number of “R” symbols in a row on the screen starting with the leftmost. The largest payoff occurs if all 10 symbols come up “R”.
The R and D symbols are generated by a mechanistic system that always generates R and D with 50/50 probability.
So we are now observing the results as we play the game. But the “D” symbol has come up at each position of the 10 for the first 4 trials, and in the 5th trial the “R” symbol came up as the first, and we didn’t note the remaining 9 symbols. So of the first 41 individual symbol trials (10 trials per slot activation), we have only had one occurrence of an “R”, and the other 40 were the “D” symbol.
Is it not appropriate to consider that we have “design” at work here?
Of course this is not sufficiently low probability. Instead we consider a further case. We run 50 plays of the slot, and in each case we get all 10 symbols as “D”, or 500 D symbols in sequence.
Now clearly this meets the criterion of the “universal probability bound” which is equivalent to the probability of one specific symbol pattern of a string of 500 bits.
Now referring back to the Caputo example, we have clearly claimed that the sequence of 500 “D” symbols in sequence would be “specified” – have we not? And if not, why?
Thus we have met the conditions to “infer design”?
[ 31. August 2003, 22:43: Message edited by: gedanken ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 01. September 2003 02:49
I think that's an excellent example. My instinctive reaction is, "The slot machine is broken." If the slot machine was provably not broken (i.e. regularity could be excluded), I would then think, "That's really weird."
Personally, I wouldn't attribute this to design. It might be design, and I'd look for a designer (or a motive, e.g. of the casino owners), but I'd also keep looking for "regularities" (i.e. mechanisms) that would explain it.
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 01. September 2003 08:04
quote:
… was provably not broken …
(My hypothetical story continues …)
They took it back to the manufacturer’s repair shop. They tested it for days and it always worked as expected. Furthermore they went through the system operation, and the manufacturer states categorically that it cannot fail. They have proven the correctness of the operating algorithms and verified the program store. The internal operation would have shown other signatures of failure if its software had failed to work. Everyone involved is absolutely confident that the system design can’t fail.
So they put it back in service. They stand around watching as various customers play – no aberrations.
Then a week later a little gray haired lady comes to the desk and complains that she put $50 into the machine and it always came up “D” in every position.
IP: Logged
|
|
rintin scrabbleweed
Member
Member # 240
|
posted 03. September 2003 16:27
Does intelligence imply “motive”?
clearly, any sufficiently complex intelligence has the capacity for motive, but when motive subverts the autonomy of the intelect - or diminishes the honesty thereof - the level of intelligence is thereby diminished;
moreover, in a community where intellect seems so highly valued, it is ironic that the motivation of contributors remains out of reach of an otherwise wide range of subject matter. ISCID will remain on the fringe of scientific inquiry until members and contributors can openly and freely discuss motivations, thereby ensuring a clear conscience and forward progress in the field of complexity.
[ 03. September 2003, 16:38: Message edited by: rintin scrabbleweed ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 03. September 2003 21:10
Hi, Rintin, your message appears to be cryptically discussing my motives. I thought I was quite transparent in my opening and following posts.
I believe that Dr. Dembski’s “Explanatory Filter” is flawed (and some variants used by other ID adherents). However I do believe that scientific methods are indeed used to discover causes of events. Those causes could be classified as “intelligent” and “non-intelligent”. Usually the scientific methods of discovery of causes don’t generalize those causes as widely as the generalization of “intelligence” without any reference to any details whatsoever about the intelligent agent or the agent’s mode of interaction with the event.
But if those causes, discovered by scientific methods, were to be classified as “intelligent” or “non-intelligent” (if those terms are well defined), then that scientific discovery of cause would be an identification of “intelligent design”. So I don’t disagree with the concept of intelligent design or scientific detection of intelligent design, rather I am interested in what techniques would actually be scientific methods to accomplish that detection.
Now in Dr. Dembski’s presentation of portions of his new book, the epigraph is shown as:
quote:
People almost invariably arrive at their beliefs not on the basis of proof but on the basis of what they find attractive. —Blaise Pascal, The Art of Persuasion
I also believe very strongly that this epigraph is extremely important and relevant. If methods of detection of “intelligent design” are based on what the adherents want or “find attractive” rather than solid evidence from observation of nature, then intelligent design is never going to succeed in being considered “science”. This is the apparent subject of Dr. Dembski’s new book. And he is writing about the “toughest questions”.
My observation about “motive” being smuggled into examples such as to make them seem to work should not be something that is feared or disdained, rather it should be examined for relevance to the questions at hand. I suggest considering that my motives, and also Dr. Dembski’s motives are the worst rather than the best. Assume the worst, that Dr. Dembski’s presentation may have been based on what he finds attractive, and assume that mine are based on what I find attractive. Assume that Dr. Dembski wants science to embrace his logic and additionally his attraction to non-material causation. Assume that I want some sort of philosophical naturalism to result. (I’m sure that neither case is even nearly correct or so simple.)
Then with those assumptions try to forget what you personally find “attractive” and look at the logic of what is presented. See what you think is the most logical implication of what is presented.
I offer a couple of pathways that I think can lead to improvement in ID methodology – though the degree of improvement may be in question as Rex has suggested.
One pathway is to recognize that a scientifically described procedure for recognizing “intelligent design” might require further characterization of the case in question, such as aspects like “motive”. Take this route and discuss the possible aspects and implications. (You might not find the implications among those you find attractive.)
Or the other pathway is to provide better encompassing set of examples of existing ID inference procedures – examples that actually provide a range of possible failures of ID methodology and show that the EF (or other design inference procedure) does not fail to be reliable in the sense of “false positives”.
[ 03. September 2003, 21:40: Message edited by: gedanken ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 05. September 2003 08:38
A very important aspect to consider: No case of the “specification” being given independent of the event (or other criterion and classifications schemes for events) will ever be completely without any connection whatsoever to the “event”. We all recognize that the matter is one of degree.
I will digress for a moment on the subject of “degree”.
For example consider a criminal investigation (typical ID example). In a criminal investigation, the more degree of support that is given for the alleged perpetrator having motive, means, and opportunity, the greater the evidence is accepted. (I mean in general human activity, not the EF specifically.)
What Dembski is doing in the explanatory filter is to create a certain generalization of classifications.
Continuing the investigation theme, the alleged perpetrator might only be classified in some thematic way, rather than identified as a specific individual. For example thefts in general might be classified as attributed to thieves, without further identifying. Then a particular “serial killer” might be identified as a characteristic individual, a “John Doe” killer with certain characteristics, but without fully identifying the suspect by name or complete details.
The investigation proceeds to try to further the classifying of the individual, as it goes along. What is very important to remember is that the details of motive, means, and opportunity are essential in refining the investigation and making scenarios being considered have greater degree of certainty.
Applying to the ID methods, we have a generalization. In the EF, Dembski specifically generalizes so as to not even classify the “designer”, not only not classifying as specific group of individuals. Even generalizing such that the possible “designer” could be an unembodied intelligence.
Furthermore Dembski not only generalizes the “designer”, but the process of “design”. Rather than narrowing to any particular mechanism of implementation, the EF generalizes strictly to characteristics observed in the final structure caused by the “event” in question.
However this generalization is one of degree, so the lack of connection of the “specification” to the “event” is also one of degree.
Here we might add to our standard list of important characteristics worth considering, when judging the certainty of the generalized classification of the “event” (chance, necessity, design): Existence, motive, means, and opportunity! Does a designer exist that could have the additional characteristics of motive, means and opportunity in the question at hand. Probabilities of each aspect (if they could be judged) could be multiplicative in considering a possible cause. (I think that “existence” is an aspect of motive – one could, for example, say that motive, means, and opportunity are all modified by existence. But “motive” presupposes existence.)
(Ending digression.)
Now the independence issue I started with is important because the specification used in normal ID examples is one generated by humans on a human-inhabited world (specifying human designs). Clearly the “independence” is not intended to be so strictly enforced that we are not allowed to use human design considerations in judging an event that might have been of human cause. Rather the “independence” of the specification and the event is a matter of degree.
We don’t go so far as to require a complete “specification” to have been written out such that there is absolutely no ambiguity that the “event” in question falls in that category. Rather the ID method of inferring “design” seems to allow for considerable latitude in deciding whether the “specification” has actually been “given” independently of the event, in the sense that the specification’s covering the event is allowed to be a matter of considerable subjectivity. This subjectivity (and thus matter of “degree”) is grounds for confusion in the application of the Explanatory Filter (EF), confusion that allows for selective examples to be given to demonstrate it so-called “reliability”.
I think that this matter of degree is important to consider (since it can never be complete). It interacts with the other knowledge we may have (possibly “side” knowledge). So I think that when I say that one is “smuggling” in the aspect of motive into examples, I am saying something that Dembski is recognizing in his writing to some extent. (The issues is logical correctness of Dr. Dembski’s conclusion of “reliability” of the EF, based on matters of degree in this issue, not whether the issue is present in some degree in Dembski’s writings.) The greater degree that motive, for example, is smuggled into examples, the greater the degree that reliability of the EF has not been demonstrated “inductively”.
The “reliability” of “no false positives” is characterized by the following:
- Cases of non-intelligent cause must be characterized by the EF procedure showing a failure to meet the complexity-specification criterion.
Actually that’s it! – It does not matter how many cases of successful application of the EF are given. If the cause were known to be “intelligent” cause, it would not matter whether the EF properly characterized the event. Either way it would not modify the “reliability” in terms of “no false positives” since having no positives at all would not conflict with “no false positives”. So we don’t even need to look at positive known results, all that is interesting to investigate are possible examples of the EF producing a positive result when the actual cause will eventually be shown to be (or could be) classified as a non-intelligent physical process.
.
Very specifically we need to recognize that all cases in which a cause is finally known (so as to test the example as an “inductive” demonstration of the EF in operation), the statements of the EF cannot be violated in strict terms. (Of course we mean these examples of failures of the EF should only occur with low probability if it is “reliable”.) For example if the event is finally known to be of physical “non-intelligent” cause, then it is very unlikely that the physical cause was extremely unlikely – precisely since it actually happened!
Therefore no EF example case of non-ID can have any actual failure of the literal terms of the EF test! So in a very real sense, the EF is tautologically “reliable” in the sense of no false positives in an absolute sense. (The very act of verifying that the case in question is one of “chance” or “necessity” is to establish that the logical conditions for the “false” evaluation of the EF!) The EF cannot ever be experimentally verified if one demands that all future information should modify the previous EF response – and that future evaluation of the EF with more complete information is taken as the test of correctness.
However the question at hand is of any practical usefulness of the EF. Can it be applied in the real world, using real-world information? This is where we must accept that information may be missing. Just as in the criminal investigation, information may be missing. Any real-world use of the EF must be evaluated based on current information as a predictor of future classification of the event in question – it cannot be considered to be modified in that result when the future information is revealed or else we are not considering it as predictive! (Science is tested in part based on its ability to be predictive.)
So how reliable is the EF, under these circumstances? Can it be improved by considering “motive” for example?
What I am suggesting is that the degree of independence of the event and the “specification” is usually compromised by side knowledge of “motive”. Examples given are actually selected such that we have knowledge of “motive” to distinguish the cases as realistic ones in which other methods of comparative analysis of causation would produce the same results.
This is where my example given above is a good one to tease out the reliability of the EF. It is a simulated real-world case of application of the EF. But it shows that we look to motive to actually tease out the question of whether an intelligent agent was responsible. So far, all the analysis has shown a match for the EF conditions, so by the EF procedure we must conclude “design”.
Can anyone give examples that do not have a comparative use of “motive” in our normal evaluation of whether the cause included “design”?
[ 05. September 2003, 10:33: Message edited by: gedanken ]
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 05. September 2003 10:42
Gedanken wrote:
quote:
Can anyone give examples that do not have a comparative use of “motive” in our normal evaluation of whether the cause included “design”?
It's unclear to me what you mean by "motive." But let's take a specific case:
What was the motive of the agents who constructed the massive stone heads on Easter Island?
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 05. September 2003 11:44
Thanks, Dr. Nelson, I added a little more to this in edit. I have some questions that relate to the issue.
I might comment that I take “motive” and “intelligence” both as fuzzy terms that might mean different things to different people. (Rex above has already criticized my concept of adding motive to the mix as not necessarily leading to greater objectivity.)
But I have some questions that I think are highly related, and will help in providing my own answer:
* What makes you think that those massive stone objects were “constructed” by “agents”? (I’m not saying that I disagree, of course, rather I am trying to understand your particular view.)
   (Just for comparison to the Easter Island faces, and the question of what makes them appear to be “designed”.)
* Where does an independently given “specification” for those heads exist? Can you formulate that “specification” and demonstrate where it exists, given independently of the event (of their “construction”)? (Since the specification cannot be given before the event, at least I don’t think it can be found, one must show independence by some other means.)
* Do you not agree that a conventional “comparative” approach works in analysis of the cause of the Easter Island faces? (I am not in this question disagreeing with the ID eliminative approach per se, rather I am trying to establish that a normal scientific approach in the case of the Easter Island figures is non-eliminative.)
Later after these questions I will try to deal with the relation to “motive” in this case. After thinking about this [returning to edit], I would think that the agents had some reason to wish to have people perceive a human image – reason enough to go to a lot of work. But this does not in itself help much in the issue I will want to get at, as this answer does not get at the EF distinction directly. We will need to construct some modified examples to see the importance, but I think that answering my questions above will help in illuminating those issues.
And I would like to point out something I said above – that tests for the “reliability” of the EF will not be changed in any way by examples in which we all agree there was “design” as a proper classification of the cause that is finally considered to be known. Reliability in terms of “no false positives” is not affected by any number of positive known causes, it is only affected by results that we must consider to in fact require a negative response from the EF, though initially a possible positive result is given. (Remember that it does not change the reliability of the EF if we agree that the cause was in fact intelligent agency, because “no false positives” is not changed even if the EF produces no positives, or if it produces a lot, in cases that are considered to be “in fact” of intelligent agency. Since I believe we shall all agree that massive stone “heads” on Easter Island were of intelligent “design,” it cannot possibly directly itself form the basis of a failure of reliability of the EF.)
Because any “inductive” demonstration of correctness of a procedure needs to consider likely cases in which it may fail, it is not sufficient to just find a few in which it succeeds. It may succeed for the wrong reasons, for example. And the cases used to demonstrate may be chosen (by design or subconscious process) so as to not have the character that is likely to make the test fail. So we really have to identify the kind of case in which the EF is most likely to fail and construct or test on that kind. That kind of case is one in which there is apparent complexity that also has apparent “specification” applicable, yet only after considerable time of analysis (perhaps years) we find the actual cause is a “chance” or “necessity” process. (Remember that the EF is tautologically “reliable” if we allow information learned later as a modification of the test inputs. A test of the EF could take presently available information as a prediction of future decision on cause of the event.)
However we may find it interesting if “motive” is considered in differentiating cases like the one given, in which some are considered to be known as intelligent agent cause, and some known as not. So I am not disagreeing that this can form the basis of an interesting case for continued discussion.
[ 05. September 2003, 23:34: Message edited by: gedanken ]
IP: Logged
|
|
Algorithm
Member
Member # 896
|
posted 05. September 2003 14:11
Can science be practiced without motive?
The “rules” of inference and their most pristine application in mathematics are very different from the “laws” of physics which math so wonderfully describes. Rules can be obeyed or ignored.
The cost of ignoring the rules of inference, scientific method, and mathematics is nonfunctionality. None of the three any longer “work for us.”
Logic, science, and mathematics are all dead if motive is not objectively real. All algorithmic and computational function immediately degrades into noise. Monod’s false dichotomy of chance OR necessity cannot explain the fact of algorithmic function in nature.
Genes are linear, digital, resortable algorithms. They are successions of integrated decision-node switch-settings. Rules, not laws, must be followed to realize biofunction.
We cannot even engage in the above long discussion without presupposing the objective reality of not only choice, but motive. Nor will we be able to understand the phenomenon of highly optimized genetic algorithms independent of “motive.” These programs are the most sophisticated known to humans. They not only predate, but produced the brains and minds of Homo sapiens.
It is the philosophic naturalist’s problem to explain what we observe in molecular biology using nothing but their metaphysical commitment to “chance OR necessity.”
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 06. September 2003 00:03
Just for understanding of the issue of what kind of example needs to be presented, here is a logic table about the EF and its reliability. (It is equally applicable to any “design inference” formal method.)
Reliability of the EF (or other “design inference” procedure) with regard to “no false positives”:
code:
Initial | Actual | Reliable
Infer ID | Cause | “no false positive”?
---------|--------|---------------------
1 Not ID | Not ID | Yes
2 Not ID | Yes ID | Yes *
3 Yes ID | Not ID | NO **
4 Yes ID | Yes ID | Yes +
* Case ‘2’ was a “false negative”. But Dembski and other ID advocates say that the ID inference procedure is not claimed to be reliable at producing “positive” results – this case is irrelevant.
** We can see that the only case ‘3’ of a failure to produce “no false positives” would be a case in which the actual cause is not considered to be “ID” (after whatever analysis leads us to confidence in the final result as to consider this a valid test of the ID pattern recognition procedure). These includes case 1 and 3, but only case 3 (yes initial “ID” inference, no actual “ID”) can actually indicate a false positive.
+ Now cases of class 4 (yes initial “ID” inference, yes actual “ID”) can be interesting in that they may be close to cases of class 3 and may stimulate construction of potential class 3 test cases. However class 2/4 cases of actual “ID” cause are not directly relevant to testing the reliability in terms of no false positives.
The intention is to see if the issue of “motive” is useful in teasing out split between cases 1,3 (cases in which the actual cause was not “ID”). Can motive be one of a set of issues considered which are not strictly “eliminative”, and which improve the ID pattern recognition procedure? Do cases that are on the boarder between 1 and 3 case turn out to be differentiated in practice by considering motive as one of many factors? (And are there other useful factors as well? However this is not the major focus I had intended for this thread).
Another interesting issue is the question of how we actually arrive at a final decision that we would all agree that the “actual cause” is classifiable as an “ID” cause. This would be the question of how we distinguish cases 3 and 4. Motive may possibly be a factor that is an aspect of convincing the large majority of us that the “actual” cause is of a “design” class. Usually we would need to be considering additional factors, because if the case in question hinges specifically on the eliminative ID procedure, it cannot be used to test itself in that the test itself cannot be used to give the “actual cause” answer by which the test itself is verified. This may be of interest in the figures on Easter Island, for example.
[ 07. September 2003, 02:29: Message edited by: gedanken ]
IP: Logged
|
|
|
Printer-friendly view of this topic