Author:
Wade A. Tisthammer
|
[
Next Thread |
Previous Thread |
Next Message |
Previous Message
]
Date Posted: 04/10/05 10:28pm
In reply to:
Damoclese
's message, "heh" on 04/10/05 6:44pm
>>And note this can carry some problems with organic
>>evolution. After all, we have seen human scientists
>>make DNA, RNA, AMP etc. from scratch but haven't seen
>>nature (outside of the cell of course) do it; instead
>>we've got a number of nasty problems for the
>>naturalistic scenario.
>
>This, to me, simply highlights the inadequacy of the
>watch metaphor. "Making" RNA or DNA isn't exactly
>taking raw materials like springs and coils and making
>a watch.
Perhaps not exactly like it but RNA and DNA are indeed artificially made. Indeed, we even have machines to do both.
>>Among other things, that is sometimes the case yes.
>>But then what does that say about some aspects of life
>>(remember what I said about abiogenesis)?
>
>It says we make attributions based on our experience.
>Life is unique in that we have no end points to
>compare it to.
That’s not quite true. Many analogies have been made regarding the complexity of life and machines (read Michael Behe’s Darwin’s Black Box and you’ll see what I mean). And if we do attribute design based on experience, what happens when our experience tells us humans can artificially create life but no known natural causes can? And to some extent this is already true e.g. DNA.
>>And even so, we still
>>don't have any naturalistic means to overcome
>>the known chemical problems and naturally produce
>>life.
>
>So what? Design must be the answer?
I’m sure you’ve heard of something called inference to the best explanation. We have an old paradigm with some unresolved problems. There is a new paradigm that solves and even predicts those problems. Or consider a little down the road when there are known artificial causes to produce life but no known natural ones (still unresolved problems with the old paradigm). Wouldn’t ID at least be a reasonable scientific theory? Or is it something to be avoided at all (evidential) costs?
>>True, but the point is the same. The premise of
>>explanatory filter is that event E began to exist
>>(which is true by definition) and the inputs are
>>correct. You haven’t shown why it can’t give valid
>>outputs.
>
>It's because I don't care about it's validity; I care
>about it's truth.
Well, that’s what I’ve been talking about. IF the filter is fed accurate information, it reliably detects design. If law and chance are inadequate, design is chosen. I really don’t see why that should be so controversial.
>>What mistake? And what would it have to do with
>>psychology?
>
>Oh, I don't know... the explanatory FILTER being a way
>to make attributions about the nature of
>something...seems like that MIGHT fall under
>psychology
In this case we’re dealing with science, e.g. statisticians and cryptologists. Small probability + specification usually does the trick in practice here.
>I really don't care if Cambridge or
>Oxford published it, the simple fact of the matter is
>that the model is worthless because it doesn't mimic
>attributional thinking closely enough to be useful.
But it is used in practice all the time. If law and chance are inadequate, design is chosen. That is often how it works.
>>And that is precisely what it does when it
>>factors in the probabilities and the specification
>>criterion.
>
>The probabilities themselves are incalculable, because
>the situation leads to inferences which rely on
>intuition.
To some extent that is true (the specification criterion is more intuitive than mathematical).
>Hence, it DOESN'T take into account the
>situation as being the driving force behind the
>attribution made, specification criterion or no.
It takes the situation into account when it factors in probabilities and the specification criterion. The filter can’t possibly work without taking into account the situation.
>>And you still haven’t given any reason why
>>it can’t legitimately and validly detect design.
>
>I'm sure throwing a dart at a piece of paper with
>three options can detect design, but I'm not going to
>be wreckless enough to put it forth as a real model
>for design detection.
You can misspell reckless and denounce the model all you want, but you still haven’t given any reason why it can’t legitimately and validly detect design.
>>I may not know the mathematical details, but it seems
>>pretty evident here that design is the most likely
>>explanation. Would you really believe it happened by
>>chance?
>
>The reason you may not know the mathematical details
>is because design is an INFERENCE.
Given the title of the book (The Design Inference) I’m sure Dembski would agree with you on some level. But you didn’t answer my question. Would you really believe the precisely ordered cards happened by chance?
>>And the magician is factored into the specification
>>criterion!
>
>No, the magician isn't as far as I can tell.
By describing the pattern in advance, he was perfect for helping the specification criterion being met.
>>>I don't think it's any MORE OR LESS unlikely, as far
>>>as the math goes.
>>
>>You better check your calculations again. The odds of
>>chance choosing that order of the cards are less than
>>one in 1067.
>
>Okay. What are the odds that that order was done
>intentionally? Got any for me? No? That's because it
>isn't quantifiable. You are comparing apples and
>oranges.
Given this scenario, I’d say the odds of design being correct are 1 – 10-67. But even if I couldn’t give the precise odds, it doesn’t really matter. Why? Science has no rigorous procedure for determining to what extent the evidence confirms a theory. It has no percentages. But this does not take away the validity of the explanatory filter. Given the small probability and the specification criterion, design really is the most rational explanation. This may only be a tentative empirically based conclusion and not a proof, but science can provide nothing more.
>>Again, you
>>haven’t said anything about the explanatory filter
>>not being able to legitimately detect design,
>>i.e. if the filter selects “design” then design is the
>>most likely explanation of the event.
>
>No, that's true, I said it is inadequate as a tool for
>determining such because it isn't consistent with
>reality.
But it is consistent with reality. The filter has a pretty big tendency to work whenever it’s used.
>>Even if it
>>wasn’t able to take into account literally every
>>factor (which to some extent I think may be true),
>>you’ve said nothing about it giving unreasonable
>>answers, i.e. nothing about its actual
>validity.
>
>Well no, I didn't, because I wasn't addressing it's
>validity, but now that you mention it, the filter
>fails when it comes to things that are complex that
>ARE random. For example, if we look at the big picture
>when it comes to quantum mechanics, we can predict the
>odds of something happening on the whole. We can say
>that even within chaos certain patterns exist. How
>exactly would this filter deal with such an occurence?
With such events, it picks law/chance whenever the situation warrants.
>>As an analogy, many integration estimates don’t take
>>in every factor. They can’t. Nonetheless they can
>>make a reasonable estimate of the true value.
>>Similarly, science is far from perfect, but it can try
>>to make reasonable estimates of what the truth is.
>>You haven’t shown why the explanatory filter can’t do
>>the same.
>
>1) It's an inadequate model in light of modern
>psychological research.
Psychological research on humans is irrelevant. What matters is whether the filter yields reasonable/correct answers. And it does. Even if humans did a completely different means of estimating integrals, it wouldn’t change the fact that the mathematical estimations still work.
>2) It can't handle certain types of order because it
>assumes the world fits into three neat sorts of
>situations.
Can you name an event that doesn’t? Didn’t think so.
>>Except it is not numbers alone: you’ve
>>neglected the specification criterion.
>
>The specification criterion DRIVES the numbers in that
>it gives them meaning based on something fuzzy.
It’s unclear what you mean by “DRIVES” the numbers. The specification criterion is used after the probability has been determined.
>>You misunderstand. Supposedly, the filter is
>>empirically verified when we detect design.
>
>But it isn't because it isn't the way people form
>attributions.
I disagree. I honestly believe that people think that, for the Rosetta Stone, Stonehenge, Mount Rushmore etc. the currents of nature are not reasonably capable of producing such things.
>>Now, given the
>>scenario that I have actually presented how could
>>design not be the most reasonable inference?
>
>Because design has no odds of its own.
So if the odds of organic evolution being true are less than one in a trillion, you still wouldn’t accept design? And if asked why, you’d say, “Because design has no odds of its own”? Well, guess what. Atomic theory has no odds of its own either, i.e. no rigorous procedure to determine the percentage of its being true based on the evidence. Nonetheless, we can reasonably accept atomic theory as a reasonable inference of the empirical data. If it were somehow known that the odds of organic evolution being true were one in a trillion, design would similarly be the most reasonable inference. I don’t see how you can reasonably believe otherwise.
> You see,
>>it isn’t the explanatory filter itself that should be
>>disputed; if law and chance are not sufficient for the
>>origin of life, then design is indeed the most
>>reasonable inference.
>
>INFERENCE!!! That's quite a bit differently than
>probability
Probabilities are used in the filter nonetheless. And while it is true that, at the end of the day, design is merely an inference of the data; it doesn’t make it any less rational (e.g. robots on Pluto example). Organic evolution is an inference of the data also. Fortunately, there is a little something called inference to the best explanation.
>>Would you honestly believe
>>organic evolution even if you knew it had odds
>>of less than one in one trillion?
>
>I'd still consider it more likely than something that
>could not have odds assigned to it at ALL.
I agree wholeheartedly, but you have still not answered the question. Would you honestly believe organic evolution even if you knew it had odds of less than one in one trillion?
Yes, the filter only provides a tentative conclusion and not a proof. But as I said, science can provide nothing more. At the end of the day, the name of the game is inference to the best explanation.
[
Next Thread |
Previous Thread |
Next Message |
Previous Message
]
|
Feedback: 