|
Author
|
Topic: No False Positives and the Lust for Certainty
|
Paul A. Nelson
Member
Member # 26
|
posted 12. November 2002 06:20
Frances wrote:
quote:
I think that it is important to realize that what Wilkins and Elsberry have done is not eliminate rarefied design but shown that rarefied design and 'we don't' know' are indistinguishable.
And given the logical structure of their filter, “we don’t know” will be favored every time –- thus excluding rarefied design (i.e., causation by an unobserved intelligence). If you tell me that your detecting device cannot distinguish A and B, but that in any case the meter will always show “B,” whatever the evidence, then you’ve effectively excluded A.
Frances wrote:
quote:
And Wilkins and Elsberry do not exclude rarefied design, they merely accept that rarefied design and ignorance are two concepts which cannot be resolved using the EF approach.
Are we reading the same paper? Wilkins and Elsberry state plainly that “no rarefied design inference is needed” (p. 722) and “why are we ever forced to a rarefied design conclusion?” (p. 721). They aren’t saying, “given the right evidence, we would infer rarefied design” or “here’s the correct way to infer the action of an unobserved intelligence.” Rather, they are simply updating Laplace’s famous quip to Napoleon: “Sir, I have no need of that hypothesis.”
Nothing in the Wilkins & Elsberry paper allows one to discover if life was caused by an unobserved intelligence.
(BTW, I heard back from Wesley – he’s travelling at the moment and can’t join the thread, but may later.)
No other probabilistic framework, incidentally, avoids the problem of incomplete knowledge. A Bayesian or a likelihood advocate (e.g., Sober) will make mistakes just like any other fallible reasoner. If what worries Wilkins and Elsberry, and you, is the possibility of false positives (error), then no model of probabilistic reasoning is going to relieve that worry. Once one begins to fear him, the God-of-the-Gaps will haunt one’s existence forever.
I’ll be leaving this and other ISCID threads for a few days, to travel with Bill Dembski to Hillsdale College for their ongoing design symposium.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 12. November 2002 13:03
Paul seems to object that Wilkins and Elsberry have correctly pointed out that rarefied design and 'we don't know' cannot be separated through mere elimination.
When they state that no rarefied design is needed, it is because the rarefied design node and 'we don't know' can be merged.
Paul seems to be focusing on the fact that Wilkins and Elsberry's filter may not be able to identify unobserved ID.
Bingo, and neither can Dembski's filter. Only by preferentially treating ID hypothesis can Dembski force an inference which cannot be distinguished from 'we don't know' but is called nevertheless ID.
It seems that the philosophical preferences of the filter designer have led to a filter which will preferentially treat ID. But a more skeptic and defensible imho filter would be one that admits our ignorance and requires positive evidence before we jump to such presumptions of intelligent design where none can be observed directly.
I agree with Paul that if one believes that there MUST be rarefied ID and that science must be forced to recognize/identify it that Dembski's filter which preferentially treats the ID hypothesis will seem to be easier acceptable than a more realistic approach which requires one to explain why one infers ID for something which cannot be distinguished from ignorance. Of course 'given the right evidence' rarefied design would cease to be rarefied. That's the whole concept of rarefied design. Rarefied design does not allow for inductive generalizations.
From a theological perspective the 'God of the gaps' argument is a dangerous one. If new data will disprove the ID hypothesis it will further undermine an issue of faith. Somehow the idea that science can prove/disprove a God seems to be unpallatable to me, from a theological as well as a scientific viewpoint.
And while no probabilistic approach can overcome incomplete knowledge, Bayesian approaches and comparison of various hypotheses can identify the BEST so FAR. It seems ironic that we should accept a design inference hypothesis which does not allow it to be compared to other hypotheses. In fact, there is no excuse to claim that low probabilities for chance/regularity hypotheses is thus evidence of a high probability for intelligent design. In fact, only after non ad hoc ID hypotheses have been proposed can one determine which hypothesis is the most likely one given our present knowledge.
So to recap, Wilkins and Elsberry's filter seem to correctly identify the same (limited set of) intelligent design examples of Dembski's design inference but the filter deals more appropriately with our ignorance and does not preferentially treat the ID hypothesis to force an ID inference which cannot be distinguished from our ignorance. Thus while the filter will allow for the same successes as Dembski's filter in cases of known ID, it will reduce the likelihood of false positives and give no preferential treatment for any particular hypothesis. In fact it recognizes that rarefied design is as much a non-hypothesis as 'we don't know'.
I can understand why people may be convinced by their faith that there must be an (un)observed intelligent designer but that does not mean that we should accept this as self evident or even that science can deal in a meaningul manner with (un)observed intelligent designers (rarefied design).
Which does not mean that rarefied design cannot become ordinary design but that would require hard work, proposing alternative non ad hoc ID hypotheses. And that's what I am proposing to be the direction ID should be taking.
IP: Logged
|
|
Daniel Edington
Member
Member # 421
|
posted 12. November 2002 22:39
Dr. Dembski,
I have some quick questions.
quote:
Briefly, the claim that specified complexity is a reliable marker of design means that if an item genuinely instantiates specified complexity, then it was designed. As I argue and continue to maintain, no counterexamples to this claim are known.
You make your point clear enough, if an item genuinely instantiates specified complexity, then it was designed. My understanding is that this method of detecting design through specified complexity is based on probabilities, basically items with low probabilities are said to exhibit this property of specified complexity. Is my understanding correct?
Here are the questions:
1.) Does this method recognize the possibility of events of low probability that exhibit complexity, but not specificity?
2.) Does the method recognize the possibility that designed objects might exist that would yield a false negative? Another way to put this is to say that the object would be determined to not be designed, when in fact it is.
3.) Does this method recognize the possibility of designed objects that exhibit specificity, but not complexity?
Dan
[ 12. November 2002, 22:46: Message edited by: Daniel Edington ]
IP: Logged
|
|
Micah Sparacio
Member
Member # 6
|
posted 13. November 2002 07:54
Dan,
Quick reply.
1.) Does this method recognize the possibility of events of low probability that exhibit complexity, but not specificity?
Yes, flip a coin 10000 times. The event will exhibit low probability but no specificity. Dembski deals with this in No Free Lunch.
2.) Does the method recognize the possibility that designed objects might exist that would yield a false negative? Another way to put this is to say that the object would be determined to not be designed, when in fact it is.
Yes, Dembski readily acknowledges that some designed events will be undetected (especially when the UPB is set at 500 bits). Designed events can have both (or either) low complexity and low specificity. Dembski deals with this in No Free Lunch.
3.) Does this method recognize the possibility of designed objects that exhibit specificity, but not complexity?
Again, yes. Consider a situation where we randomly drop scrabble pieces on a scrabble board and we just happen to form the word "the" Dembski deals with this in No Free Lunch.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 13. November 2002 12:41
Is CSI a reliable marker of ID? It seems contradictory to claim that algorithms cannot generate CSI since the probability of them generating such information is close to 1 thus high probability or low complexity. But why does the same argument not apply to ID? ID certainly can generate the same information with the high probability, or in other words the observation E given the hypothesis H of Intelligent Design is highly likely. Thus the complexity once again is low. Perhaps a better observation may be that neither Intelligent Design nor algorithms can generate CSI because of the high probabilities involved. Does this mean that when we see CSI that this is evidence of ID? As I have argued, no. CSI at most is evidence of our ignorance since we cannot formulate the correct hypothesis to explain it.
Thus I would argue that CSI is a reliable marker of 'we don't know' rather than of ID.
The moderator has argued that my claims are 'the same criticisms' but the argument, although leading to the same conclusion, is different. Based on CSI, I argue that neither ID nor regularities can generate CSI. This is an important observation in the sense that CSI is argued by Dembski to be a reliable indicator of design. I am arguing that in the instances of ID which have been considered and for which we have known ID, the amount of CSI is close to zero for the same reason as argued for algorithms, namely the probability of the event is close to 1 given the ID hypothesis, thus reducing the complexity to close to zero.
Unless others have argued this on ISCID, this seems to be the first time this argument is proposed. In short: if algorithms can reduce the CSI to close to zero because the probability of the event is close to 1, why does the same not apply to ID hypotheses? Let's for the moment drop the argument that high improbability/high complexity may be a more reliable indicator of something else. It seems important for the design inference that we determine indeed if CSI can be generated and if so if there is a difference between CSI from algorithms and from ID.
[ 13. November 2002, 13:28: Message edited by: Frances ]
IP: Logged
|
|
Moderator
Administrator
Member # 1
|
posted 13. November 2002 13:00
Frances,
I warned you earlier about abusing your time at ISCID. This is my absolute last warning. Stop spamming us with your doctrine. By repeating the same criticisms over and over, you make it clear that you are here with an agenda. At least five different brainstorms members have independently indicated that they are getting tired of your methods. I'm getting tired of your methods.
Though I was just about to throw the switch and ban you for 1 month, I've held back. Make me glad that I did. To do so, you're going to need to do a little more than chant the same refrain. Think a little before you pull the trigger to post next time.
IP: Logged
|
|
Daniel Edington
Member
Member # 421
|
posted 13. November 2002 17:05
Micah,
Thanks for the quick reply, however I am not entirely sure you answered my questions.
quote:
1.) Does this method recognize the possibility of events of low probability that exhibit complexity, but not specificity?
Yes, flip a coin 10000 times. The event will exhibit low probability but no specificity. Dembski deals with this in No Free Lunch.
So we can safely say highly complex objects can arise that are not specified. What if one randomly tosses the same 10000 coins and the results actually correspond to some known binary pattern (the equivalent idea of randomly dropping scrabble pieces and spelling the word “the”.) Such a result would be just as unlikely as any other pattern. Would such a pattern then be specified complexity? What of the random tosses resulted in the same known pattern, except a small percentage of the digits were wrong, would it still be specified complexity?
quote:
2.) Does the method recognize the possibility that designed objects might exist that would yield a false negative? Another way to put this is to say that the object would be determined to not be designed, when in fact it is.
Yes, Dembski readily acknowledges that some designed events will be undetected (especially when the UPB is set at 500 bits). Designed events can have both (or either) low complexity and low specificity. Dembski deals with this in No Free Lunch.
No I was thinking of highly complex objects resulting in false negatives. Your answer also raises another good question. I understand what 500 bits of information is where my computer is concerned. However, I am not sure how much 500 bit of information is with respects to other systems. How much information is contained in a typical protein for example? Would an object such as a snowflake be considered complex and how much information would it contain?
quote:
3.) Does this method recognize the possibility of designed objects that exhibit specificity, but not complexity?
Again, yes. Consider a situation where we randomly drop scrabble pieces on a scrabble board and we just happen to form the word "the" Dembski deals with this in No Free Lunch.
If it is a chance event how can you say it is designed? It may in fact be specified, but it would be after the fact wouldn’t it?
Dan
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 14. November 2002 23:31
I have been looking into more detail into the CSI argument and I am confused about when one calculates CSI relative to the uniform probability hypothesis and when one calculates it relative to the actual hypothesis? I was under the impression that CSI was calculated relative to the hypothesis.
David Wolpert has reviewed No Free Lunch and reached similar conclusions about the need for precise mathematical definitions.
I find Wolpert's feedback quite interesting since Wolpert and Macready are the authors of the infamous No Free Lunch theorem papers.
I think that the ID inference could benefit from some more precise mathematical definitions.
An author by the name Erik has attempted a more formal definition of the design inference. Could we use his definitions to tighten up the mathematical foundation of the design inference?
For instance I have raised the question if CSI iscalculatedwrt the uniform probability function? or wrt the actual hypotheses?
Some of Dembski's earlier work in NFL suggests the uniform probability function but in his response to Wein he suggests that [url=www.designinference.com/documents/ 05.02.resp_to_wein.htm]Consequently, for complexity to obtain, the probability of the specified event in question must be less than the universal probability bound with respect to all the probability distributions[/url]
This however means that once a hypothesis has been found which explains the data that the CSI has reduced to a number close to zero (probability goes to one, log of probability goes to zero). But if this is the case, any hypothesis which explains the data will have its CSI be reduced to zero.
What I am doing wrong here?
How can we define these important terms in a manner which make them accessible to rigorous testing? After all the next step of the ID should be developing a scientific research program? and the detectability problem seems to rank high on the list.
[ 14. November 2002, 23:34: Message edited by: Frances ]
IP: Logged
|
|
John Wilkins
Member
Member # 418
|
posted 27. November 2002 20:18
I'm finally back from the wilds of New Zealand, and able to connect this beast to the internet, so here is my response.
Paul Nelson has questioned whether there are any conditions under which I would accept a rarified design inference (I cannot speak for Wes, and he can, so let him). This is a good question. Clearly there would be some situations where I would rationally have to accept such an inference. A finger descending out of the clouds in Monty Python style and tapping me on the shoulder might do the trick (in preference to the inference that I was hallucinating or mad, perhaps, if the other information pertaining made that unlikely to me).
However, this is not the issue with Dembskian/Paleyesque natural theology. That is not about what would rationally force a rarified design inference for an individual (which is, in any case, going to depend a lot on the relative weights of the individual's beliefs in their own rational scheme). Instead, natural theology is about how much knowledge of the empirical world can tell us about the nature and existence of a designer of that world. In short, the issue is what science can tell us, and indeed does tell us, about God. Nothing in the modern design movement is anything other than such Paleyesque claims, and they would be just as familiar to John Ray.
Let us begin by asking how science tells us anything at all. In my view, science is the recognition of patterns in data, and the generation of models that are adequate to delivering those patterns as explanation. The information in science, the "signal" from the physical world, is the information of measurement - Fisher Information, AKA the Cramer-Rao Bound (which is, roughly, where the second derivative of the estimate of the accuracy of a measurement is zero). In my view, science is induction from data, and the models retain the information content of the measurements just to the extent they are accurate. (Note: induction may not be a justification of models, but it sure as hell is the way we gather our data together so we can make reliable inferences; still, let's not open that can of undergraduate Humean worms.) The information content of a scientific explanation is just the preserved accuracy of the data in the model.
Anything that we know through science we know from empirical data. So a design inference has to be not only consonant with data, but licensed by the patterns that exist in the data. To be achievable, we need to understand (that is, have a model of) design and designers. Consider three kinds of designer: known ordinary (type A), unknown ordinary (type B) and unknown rarified (type C).
We already know how type A designers work - we are they. We can observe them, experiment with them, and they can express their intent. From this, we can develop (and in folk psychology we have in-built) a model of design. We can assess when a design goal has been successfully met or when it fails. We can compare two designs and tell which is better according to the stated, and sometimes the unstated, goals and criteria for success. Moreover, we know the laws of physics and materials that apply to designs, and so we know what may have been intended when the intent and criteria are hidden to us (in the case of ancient artifacts). Notice, though, that the less culturally like us human designers are, the less obvious the design intent is. With a lot of stone tools from the paleolithic, we can only guess what they were for. And that is with humans, whose ways are as our ways.
What of designers whose ways are not our ways? What of type B designers? I'm not talking about minimalist architects here, although they are about as alien as one can be with a human anatomy, I'm talking about non-human designers. What do we know about their designs? This is really a question of what we would be able to recognise as design. For terrestrial organisms we are more or less related. We can recognise design among mammals, and to a lesser extent birds, although the less like us the more the tendency to anthropomorphise (the biological equivalent of eisegetical interpretation). Still, if a crow in Japan puts a nut under the tyres of a car stopped at a stop light, and returns to gather the contents after it is cracked, then that is design; because we know that crows eat, that they have problem solving skills, that shells are hard, and that cars are heavy. In short, it is how we would do it, if we were in their place.
But now what about something that is very unlike us? What about type B designers that are discovered through SETI, to use one of Dembski's favourite examples (although I shall not make any inferences based on the novel or film Contact, as it is fiction, and philosophers are altogether too fond of using Gedankenexperiments as if they proved anything)? Well, we still know the properties of the physical world, and we might expect that they would use the same maths as we do (although if the Wignerian explanation of why maths is useful is correct, there is no guarantee of this - they may evolve a maths quite unlike anything we can envisage, because they interact with the physical world very differently to us; still, I would expect it to be something we could describe in some ideal mathematics). We might therefore be able to say of a certain signal that it is designed because the physical world only delivers such signals when there are designers in play. Exactly how this might work I am not sure; and neither is anyone else. Certainly not Jodie Foster, for all her undoubted virtues. So much for SETI - it tells us nothing until we actually have cases in hand, really, other than how our expectations might work. But again we can say of some fictitious case that that is how we would do it, if we were in their place (because, with the fictitious example, we are - it is only in our imaginations).
Now, here's the Big One: what about a type C designer? This kind of designer is not bound by our logics, nor by our physics, nor by our cognitive propensities. Any pattern of data generated by the agency of a type C designer is either going to be indistinguishable from the patterns generated by ordinary designers, and hence something we will interpret as an ordinary (lawlike) process, or it will be indistinguishable from a random process - in effect we will not have any reason to gather into a single class all the measurements that are due to this supernatural intervention (let us be frank about this - if we are talking about pansperming aliens, they fall under type B). How could we do this without begging the question?
Consider Darwin's reaction when he encountered the flora and fauna of my own country Australia. His immediate reaction, as recorded in his diary and the Journal of Researches (well before the transmutationist theory) was that it seemed like "two different Creators were at work" - one for the Eurasian and American animals and one for the Australian. But why stop there? Why not posit that there is a designer for humans, and a designer each for Cryptosporidia, Yersinia pestis (plague), smallpox, etc? Why not separate designers for the predator and the prey in each case? Each is "designed" for a suite of tasks, and these tasks conflict. There is no way you can extrapolate from the adaptations of a gazelle to escape lions and cheetahs, and the adaptations of lions and cheetahs to bring down gazelles, to a single designer for both sides. On purely empirical grounds, you must infer a separate designer for each organism (which rather undercuts the rationale for the designer hypothesis in the first place, I think).
But we routinely hear from the design proponents that there is a designer required to account for all life. This is not something that one gets out of the data; how does one arrive at it? Obviously, from the prior metaphysical commitments of the design proponent. Why not have two designers, like the manicheeists? Why not have a Good designer and an Evil designer? It seems to me as good or better a hypothesis - at least it is consonant with more data. The obvious answer is that it is only on the basis of monotheistic doctrine that we can rule it out. Note that the multiple designer hypothesis is a henotheist doctrine at best, and out-polytheises Hinduism at worst. I rather like it - in Pratchettian "small gods" terms, it means there is a god for every adaptation. A god of sickle cell anemia, for example... I might
The plain truth here is that the data does not support a design inference of rarified design, and it is barely consistent with a type B ordinary designer, and in neither case is it necessary for life to be explained in terms of design - we have a perfectly adequate model (which antievolutionists and philosophers misleadingly call the "Darwinian" model, which is a number of distinct theories not all of which are common to all who consider themselves evolutionists).
But there is another possible designer - the type D designer; this is the unknowable rarified designer. This designer cannot be established in terms of empirical data for the reasons the type C designer cannot be - we have no inductive patterns that lead us to make that inference more likely over "we do not know", as Wesley and I argued. But the type D designer doesn't need this because this designer is not knowable a posteriori. This designer is the designer of revelation, and revelation is what makes some patterns of empirical data significant; it is what puts the outcomes of supernatural agency into a single class. If the designer you seek is one that reveals itself in non-empirical ways (i.e., outside the purview of any science as such) then I can have no possible objection; and if that revelation is enough for me, then I will accept that inference. But I shall not, now or then, call it science or think that the empirical world is what licenses that inference.
Incidentally, on a literal reading, that makes Paul's "without excuse" claim in Romans false. Everyone who is not the recipient of suitably convincing revelation is excused. But it seems to me that Christians, who do not wish to adopt a dualism like Mani's, must not argue in favour of a natural religion parallel to or supporting revelation - this has to be a heresy. I cannot speak for Jews or Muslims or other religions, never having been or studied them in detail. This also undercuts the traditional form of deism. There can be no empirical theism.
That should provide you enough to label me in some manner. However, I think that it clarifies Wes' and my argument that rarified design arguments can never be distinguished from ordinary lawlike explanations or chance; in fact, Paul, if you think so, then there is something wrong with your arguments and model, in my opinion.
IP: Logged
|
|
Mark Szlazak
Member
Member # 391
|
posted 28. November 2002 01:14
John Wilkins has just sketched out an argument against 1) the over generalization of the the design detection algorithm of Dr. Dembski AND 2)neo-Darwinism. I like it, and plan to run with it after the holidays.
I don't know if it's just me but for some reason this specific detection procedure reminds me of something that the RAND corparation used in the 1990's when developing health care guidelines in the U.S. Remember the techniques of meta-analysis. Ways to combine different studies to TRY to get a more objective review. It has it's limitations and can be abused ... and I was never really fond of it for various reasons.
Maybe I'm hallucinating but that's how it seems to me right now. I'll see if this looks like a bit of a turkey after I've had my turkey.
[ 29. November 2002, 01:47: Message edited by: Mark Szlazak ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 28. November 2002 02:14
May as inquire as to the nature of the 'against neo-Darwinism' claim in your posting? What exactly led you to such conclusions?
IP: Logged
|
|
John Wilkins
Member
Member # 418
|
posted 28. November 2002 03:13
I'll bite. How is my argument against neo-Darwinism?
IP: Logged
|
|
Mark Szlazak
Member
Member # 391
|
posted 03. December 2002 20:57
John, I'm sorry it's more an issue with the language you used. Materialists typically deny that intent, desire or feelings exists (Dennett, Churchland) or if they do then they're not causally efficacious (epiphenomanalists).
It sounded like you were coming from a position of an idealist, dualist or panpsychist. From these perspectives a series of disconnected intensions over a period of time or all at once could produce a result that appears to have been done with one intension in mind but wasn't. Happens in daily life all the time. Some extreme examples are the conspiracy theorists that make a living claiming some intent or hidden agenda was behind it all when there really wasn't any.
Entropic accounts don't necessarily tract causation or knowledge well and as others have said, a validated causal account is needed.
[ 04. December 2002, 14:03: Message edited by: Mark Szlazak ]
IP: Logged
|
|
Paul A. Nelson
Member
Member # 26
|
posted 04. December 2002 10:24
Hi John,
I'm back from Thanksgiving vacation. Thanks for your thoughtful post.
Do the following two points summarize your position correctly?
1. Rarefied design is knowable (i.e., can be inferred in science), given sufficient evidence.
2. At the moment, however, the evidence does not support rarefied design (i.e., the data are better explained by other theories).
Let me know if I've misunderstood you.
[ 04. December 2002, 10:26: Message edited by: Paul A. Nelson ]
IP: Logged
|
|
John Wilkins
Member
Member # 418
|
posted 08. December 2002 16:07
Paul Nelson asked me
quote:
Do the following two points summarize your position correctly?
1. Rarefied design is knowable (i.e., can be inferred in science), given
sufficient evidence.
A rational conclusion for an individual given some set of assumptions and a certain kind of evidence (which will be different for each individual). But not, so far as I can tell, a rational conclusion for science per se to draw, because of the way science handles information (i.e., empirical data). Science is the process of drawing the broadest possible conclusions warranted by the information "contained" in the data. My point is that although I may be personally forced to conclude that something is rarified design, that is not and can never be scientific, because the evidence is not interpretable in science. Nothing can license that conclusion as a scientific result, because in science such interpretations are inadmissible. As soon as you do make such claims, then you are exceeding the information content of the data. And also, as we said in the paper, such inferences would stop science dead in its tracks.
quote:
2. At the moment, however, the evidence does not support rarefied design
(i.e., the data are better explained by other theories).
Not necessarily. We can rest satisfied with a nescient conclusion (thanks for that term, by the way, I shall use it a lot) in science. Science does not need to have a theory or explanation for everything; well, not yet anyway. I may personally need to answer all these questions for my own benefit (as it happens I personally do not, for reasons I will discuss in private with you if you like), but that is not a scientific inference. At the most, it is an inference that relies on the facts being right (i.e., the information content of the data must not be contradicted).
Let me give an analogy. I am something of a moral realist (strictly, I believe that statements in moral language are "true or false" in a moral language game independently of factual statements). Hence, I do not infer my moral beliefs from factual statements. However, in so far as a moral precept relies on things being thus and so, those factual statements have to be correct - I cannot say, "All humans are required to fly to Mecca unaided", since humans cannot fly unaided no matter what. I cannot say either that humans must synthesise their own vitamin C, because we cannot do that. "Ought" implies "can" as Hume said, or more correctly a moral obligation relies on our ability to carry it out.
Likewise, the science in any metaphysical story must be correct, but no metaphysical story is science. The difference is that rational inference in an individual is relative to the commitments of an individual, but in science there are no such shared commitments beyond an attention to empirical measurement, and the construction of models to generalise within a domain of explanation.
quote:
Let me know if I've misunderstood you.
You have, but that may be more due to my inability to express myself than anything else.
IP: Logged
|
|
|
Printer-friendly view of this topic