« Blue-ray and HD-DVD: wave of the future, or already washed up? | Main | And now for something Purely Commercial... »

ID debate, and who's in charge?

Foreword: Dean Esmay writes about A Voice of Sanity on Intelligent Design. After I finished writing this, I was surprised by my conclusions. As opposed to Dean, I'm agnostic, and believe there are "things on Heaven and Earth, undreamt-of in your philosophy." On the other hand, I still insist on scientific rigor when processing ideas.

So where does that leave me?

-Scott Kirwins says "Natural selection and Evolution has whethered[sic] 150 years of attacks and grown stronger with time."

So you've got the debate between steady-state evolution vs. catastrophic evolution all sussed out, Scooter? You'd better write Science and let them know! :)

Point being that there are many areas of uncertainty within the overall theory. Painting the basic theory as unassailable begs the question. We may find encounter data in the next fifty to one hundred years which calls at least some of the current assumptions into question. Einstein developed a theory which strongly altered Newtownian physics, but not until new data was introduced.

-JDS brings up various philosophical beliefs, 99% of which are utter twaddle. Compare the practical results of the last 2,000 years of philosophy, as opposed to the practical results of the last 200 years of science. Real science is easy to define: if you can't measure it, it ain't science. If it is science, you use mathematics for analysis.

Of course, this means that many areas called science, aren't, including psychology, psychiatry, sociology, and so on. You can call them studies, disciplines, or even areas of expertise. They just aren't science. JDS, alas, drops the ball with the claim "the value one places on science is, in itself, a philosophical view." Untrue.

Real science is -as I said above- measurable, repeatable, and subject to mathematical analysis. Real science (in fact) deals in objective, not subjective concepts. This is (as has been mentioned previously) why science cannot address the issue of God. There are no "God fossils" lying around, and He has been recently uncharitible in providing a convenient miracle to confound the skeptics. ;-) We are therefore forced to rely upon unpleasant facts, logic, and statistical analysis.

I would, therefore, like to take issue with Martin's remark that 'Scientists use statistics all the time to say: "This event is so statistically unlikely that we can dismiss it as impossible."'

That turns out not to be the case. An actual statistical analysis gives only one of two results: reject, or cannot accept or reject. I have to add that the latter is sometime referred to merely as "not reject," but I find that unclear, with an implication that "not reject" is equal to accept.

It works like this: one tests a hypothesis (say: smoking causes lung cancer) by creating a null hypothesis opposite to the starting hypothesis. In this case the null hypothesis would be "there is no connection between smoking and lung cancer." The next step would be in analyzing the statistics, which (the results have been established for years) show that there is a relation between smoking and lung cancer. We then reject the null hypothesis. This allows us to accept the original hypothesis, at least to the extent that we use further analytical techniques to further test it.

Please note that you can't use this approach to "prove" a hypothesis; the best you can do is not reject a hypothesis. You can view falsification as winnowing the scientific wheat from the inaccurate chaff.

I suppose I should point out here that a 90% level of confidence doesn't mean the research is 90% confident; rather it means that 90% of the time, the actual (as opposed to statistically estimated) value will fall within a specific interval. To take a common example, an approval rating of 35% [+-3%] for the president, with an established confidence level of 90%, means that 90% of the time the actual approval rating of the entire country will fall between 32% and 38%.

What most folks miss is the immediate corollary: 10% the actual value will fall outside of that range. This means that 10% of the time one (incorrectly) rejects the null hypothesis. For the above example, this would have meant that there was no relationship between smoking and cancer, despite the data indications.

This is known as a "Type I error:" incorrectly rejecting the null hypothesis. A "Type II error" is when you fail to reject a null hypothesis. Again, note that a Type II error doesn't mean one has incorrectly proved a hypothesis, but merely that you failed to reject it. It is, in other words, not evidence to the contrary.

I just came across an interesting analogy while double-checking my memory: this is similar a court decision, wherein "guilty" is the definite rejection of a false hypothesis, and "not guilty" is failure to reject. Most people equate "not guilty" with "innocent," which are both legally and semantically different. One may fail to convict an actual criminal due to lack of evidence. One may fail to reject a false hypothesis for the lack of data.

What this boils down to is that -scientifically speaking- one may disprove a false hypotheses, but not positively prove a true hypothesis.

So those who say that any scientific theory -evolution, for example- has been "proven" are incorrect in their statement. It would be more accurate to say that the theory in question accurately explains the known data. Also note that I do not say that they are wrong, but rather inaccurate. A proper respect for science tends to instill respect for semantics as well. :)

This common confusion about statistics is also reflected in the common confusion about probability, as the study of any state with a lottery can attest. One of the more popular (bad) arguments is "argument by unlikelyhood."

You can even see this in political discussions, wherein one disputant makes the claim that event X was so improbable that it is only logical to assume that conspiracy/criminality/take your pick is the only possible answer.

The first problem with this is that these arguments never establish actual probability distributions before the fact, which begs the question. How unlikely is an event? Say someone rolls two dice 108 times. Furthermore, the roller gets "snake eyes" (two 1's) three times. An observer then immediately claims the dice are loaded, since there's such a low chance of rolling two 1's, which is non-quantitatively true. But an objective examination of the probability distributions shows that -in 108 throws of a pair of honest dice- one should expect precisely three sets of "snake eyes." This is a simple example. Generally, one must specify the probabilities before an event, not after.

Another area of confusion is trying to estimate after-the-fact probabilities. Our example: every US dollar bill has a serial number consisting of a letter, eight digits, and another letter, e.g. B11895196B. What are the odds of my having that particular dollar bill?

(For those completely unfamiliar with probability, the probability of a combination of statistically independent events is equal to the product of the separate events. In the above example, the probability of rolling "snake eyes" is 1/6 x 1/6 or 1/36. The probability of rolling three 1's with three dice is 1/6 x 1/6 x 1/6, and so on. The probability of picking out two B's out of a bag of SCRABBLE tiles (assuming a single tile for each letter, and replacing the tile between picks) is 1/26 x 1/26, or 1/676, which is about 0.14%)

Someone who knows a bit about basic probability would reply "well, that's a series of independent probabilities, so you multiply 1/26 x 1/10 (6 times) x 1/26, and get 1/(26 * 1,000,000 * 26), or 1/676,000,000. Thus our math wizard answers that the author has a one out of nearly seven hundred million chance of having that dollar bill in hand. A very unlikely event, indeed!

There's only one problem with that: the actual answer is precisely 1.0.

How do you resolve this apparent paradox? Simple: the event has already occurred. That is, I already had the dollar bill in hand when I phrased the question! (Literally. I took one out of my wallet to get the example serial number.)

In other words, the probability of an event which has already occurred is always 1.0, or 100%. Therefore anyone who argues that event X (having already occurred) is suspect, because the event is statistically unlikely, is providing an argument which is statistically irrelevant.

...Which -now that I think about it- seems to shoot a pretty big hole in ID. Whoops. On the other hand, I doubt this would impress them very much, as they reject the classic definition of science I presented above, in terms of "real" science is that data which can be measured, analyzed, and reproduced. If you can't use the math, don't call it science.

Actually, if you go here, and read some of the arguments presented by the Discovery people, you'll find they're hyping philosophy over science, preferring metaphysical hot air (but I redund) over objective, measurable reality. Basically, they attack real science as "positivistic," and try to vitiate the elemental concepts of measurability, observability, and falsifiability.

Instead (assuming that Steve Meyer's article is representative of the whole) they try to make ID acceptable by formulating "philosophical" definitions of science which allow the introduction of supernatural intervention via appropriate definitions.

By the nineteenth century, attempts to distinguish science from non-science had changed. No longer did demarcationists attempt to characterize science on the basis of the superior epistemic status of scientific theories; rather, they attempted to do so on the basis of the superior methods science employed to produce theories. Thus science came to be defined by reference to its method, not its content. Demarcation criteria became methodological rather than epistemological.
(emphasis added) In other words, Meyer prefers the philosophical, subjective approach -arguing about "true" vs. "false' knowledge- as opposed to the modern, objective approach, which still relies on measurement and analysis, despite what he has to say about it, such as his complaint that Newton's laws did not "explain" gravity
First, many laws are descriptive and not explanatory. Many laws describe regularities but do not explain why the regular events they describe occur. A good example of this drawn from the history of science is the universal law of gravitation, which Newton himself freely admitted did not explain but instead merely described gravitational motion. As he put it in the "General Scholium" of the second edition of the Principia, "I do not feign hypotheses" -in other words, "I offer no explanations." Insisting that science must explain by reference to "natural law" would eliminate from the domain of the properly scientific all fundamental laws of physics that describe mathematically, but do not explain, the phenomena they "cover." For the demarcationist this is a highly paradoxical and undesirable result, since much of the motivation for the demarcationist program derives from a desire to ensure that disciplines claiming to be scientific match the methodological rigor of the physical sciences. While this result might alleviate the "physics envy" of many a sociologist, it does nothing for demarcationists except defeat the very purpose of their enterprise.
In other words, Meyer insists that a natural must explain the ultimate "why" of a process, as well as the process itself.

I could go on, but basically the above article (and again, by extension the ID position) relies on philosophy and fancy tap-dancing about definitions instead of science.

True, Carl Sagan once observed that the scientific process does include at least one metaphysical concept: the world is a rational place, and operates according to physical laws, which can be derived and comprehended by human beings. So far, that assumption has withstood examination.

The odd thing about this is that I have no theoretical objection to God, I just expect His works to follow the known laws we have derived from our (aha) God-given faculties. If you check up the Roman Catholic position these days, the Pope decided a few years ago that evolution was compatible with Christianity, as long as one posited some sort of "divine spark" inherent to the speciation of homo sapiens. It's gonna be bloody difficult to test the idea, though!

A major part of the confusion in this debate is, I fear, yet more confusion about terms. This isn't really about science, or the "philosophy about science" (what the fubar is that, anyway? how can you get philosophical about the Fourth Law of Thermodynamics?) It's about control of the political entities in our lives.

In this case, it is the question of who runs the local school. Is it the local community, the county, state, or federal authorities? Jerry Pournelle has repeatedly pointed out that the best answer to this question is to reduce authority to the smallest denominator; in this case the local school board. As the good doctor has pointed out, this will no doubt result in some very silly-looking reading lists, but that's the cost of liberty. The beauty of this approach is that 90% of the worthless dolts (er, politicians and lawyers) who interfere now, can't.

This isn't a case for the courts, nor even for the state legislature, as long as the school isn't breaking any laws. Certainly it's none of D.C.'s business.

The problem is that everyone likes to talk about liberty, but they hate to let other people exercise it. Simply put, the left breaks out in hives about "God in the classroom," while the right wets their pants about "sex in the classroom." Here's a new idea: let the parents decide. It's their kids, and it's none of our business. If you're that worried about it, work on your own kids first. If you don't have any, shut the Hell up, as it's literally none of yours.

TrackBack

TrackBack URL for this entry:
http://www.thegantry.net/cgi-bin/mt/mt-tb.cgi/219

Listed below are links to weblogs that reference ID debate, and who's in charge?:

» More Discussoin from Dean's World

Casey's got some more discussion going on the ID stuff.

So does the Commissar.

[Read More]

» Intelignet Design Debate Heats UP in the Blogs from Vista On Current Events

Tuesday I linked to Dean Esmay's Post "A Voice of Sanity on Intelligent Design." Dean is a non-theist and a believer in... [Read More]

Comments (11)

Casey, thank you for providing a different (and intelligent) perspective on this issue. I must admit, I've long since almost despaired of creation / evolution / ID <strike>fights</strike> discussions in the blogosphere— they almost always resolve into an atheism versus fundamentalism <strike>folie à deux</strike> debate, with anyone who doesn't fit into one of those two categories pretty much excluded from the debate (or at least ignored within the debate) from the outset.

And I say that as a theologically traditional Presbyterian who has always accepted evolution without any problem.

As for ID, I can't get very worked up over it one way or the other, perhaps because I find it neither threatening nor plausible. You're probably right, let the teaching of ID be resolved at the level of the local school board— though I really can't work up enough of a head of steam over ID to be strongly fer or agin' any particular policy prescription. So I guess what I'm really saying is, let the chips fall where they may, and frankly, Scarlet, I don't give a damn. :-)

Casey Tompkins:

That's pretty much my position, too, Paul, which is why I was surprised by my own conclusions by the time I had worked through the statistics examples.

It was pure serendipity I finished with the example of an event which as already occured.

This article has caused further reflection (on my part) regarding just how many people view science as a dogma, or a secular religion which reveals "truth," as opposed to factually accurate models.

I agree with most of what you have to say Casey but I must disagree with some of your details.

First, it's wrong to state that sociology is not a science because it doesn't involve math. And it's wrong in two ways: first, not all science requires math--empirical observation is often enough--and second, because there's tons of math involved in sociological study. Indeed, you refer yourself to one of the most important and useful contributions that social science has made: public opinion polls. Indeed, probably one of the largest contributions that science has made to civilization over the last 100 or so years has been the massive introduction of statistics into everyday life through social science. So you're hoist with your own petard there my friend: you cannot simultaneously say sociology (or political science, or economics) are not true sciences while then using a classic example of sociological science--a presidential approval rating--as scientific. ;-)

Game Theory is another area where social science has proven quite capable of making predictions mathematically, by the way.

Also, I have to note that scientists DO in fact accept that some events are so statistically improbable as to be considered all but impossible. Just for example, it is physically possible that in their random movings about the room you're sitting in, all of the oxygen atoms could migrate to one corner, causing you to perish because the only thing to breathe near you is C02, nitrogen, hydrogen, and so on. That is a non-impossible event. The odds are so astronomically low that such an event has likely never happened to anyone in all of history, and likely never will happen to anyone in all of history. But it's non-zero.

What you seem to be suggesting, furthermore, is that if a statistically unlikely event occurs, we must assume it happened at random. So, for example, let us say that when we find your body and do an investigation, we prove that all the oxygen molecules in your room migrated to one corner, and that is why you died. Are we free to conclude that this event happened randomly, no matter how unlikely? Or should we investigate the possibility that some murderer CAUSED the oxygen to so move in order to kill you?

Let us say I show you what looks like an old stone arrowhead. I can show you why it is statistically unlikely that this rock got this exact shape purely through erosion. You can't just say, "well, it happened through erosion because its existence is proven--odds are 1.0!" No, the odds of the arrowhead existing are 100%, but the odds that the arrowhead was created by hand are infinitely higher than the odds that random erosion produced it.

Even your own experiment with the dollar bill shows us intelligence. Let us say that you did not reveal to us that you intentionally read off the numbers from that bill. What if you simply said, "here's a sample serial number I've just typed... and look, I happen to have a dollar bill here with just those numbers!" The existence of the bill is not in question, but if you tried to tell me, "that happened entirely at random," I would say, "not bloody likely you crazy BS'er."

On the whole you make good points though. ;-)

Ack. Close the I tag Casey. ;-)

Jason:

You state: 'JDS, alas, drops the ball with the claim "the value one places on science is, in itself, a philosophical view." Untrue.'

How is this untrue? You never really say. You obviously place a great deal of value on science. Others do not. That they do not is a philosophical point of view. One I do not share, but it exists, nonetheless.

You say: "Real science (in fact) deals in objective, not subjective concepts."

This is entirely the point. Not everyone (or every philosphical movement) values the objective over the subjective. To provide an example, read some of the arguments in the debate between the value of qualitative and quantitative research. Another example is Guba and Lincoln's chapter in the Handbook of Qualitative Research that discusses different philosophical paradigms.

While I don't wish to take issue with you about what "real" science is (although I agree with Dean's points), but I'm just pointing out that there are researchers out there who don't take a particularly positive view of "real" science. To put it politely, they have a more "open-minded" view of what constitutes the search and discovery of truth.

I only brought up the point to illustrate that some don't view being called "unscientific" as an insult, so you're unlikely to pursuade them by calling their views "unscientific."

The whole point of my argument was simply to state that different philosophical movements place different values on "science" as a "way of knowing," so to speak. Science is a product of philosophy, not something entirely separate from it.

John F Harvell:

(Bias alert) I happen to be one of those who would like to see ID become an accepted hypothesis.

That said, Casey, you have done an excellent job of describing my own view of science as compared to the rather dogmatic assertions I see so often in the media and in reportage of debates about evolution. In my own mind, I have also compared Darwin's theories to Newton's, in that they are both reasonable approximations to the real world, based on the available evidence. As you accurately point out, neither of these is "proven" in the absolute sense but both are useful working hypotheses under most ambient conditions.

But I would also argue that, just as Einstein went a step beyond Newton based on evidence based on new measurement technologies and concepts, there is a reasonable possibility that there is more to the origins of life than random selection. I haven't read Stephen Meyer's reference that you dislike, but I find William Dembski's statistical argument pretty persuasive. In objective terms, Dembski's argument does not say anything about God, merely suggesting a superior intelligence. That could also be construed to mean space aliens just as well. In the broad view, ID could well be just the next step in expanding on the special case that Darwins theories describe.

If you haven't had occasion to look at Dembski's work, he does maintain a website under his name where you can examine his thesis.

Casey Tompkins:

Dean's got me dead to rights on his first point. I completely forgot about a major category: natural science. I doubt Carl Linnaeus used much math when developing his taxonomy, but that's demonstrably a science.

I have to wonder if you can put Games Theory in that category? If I had continued the series in Systems Analysis from deterministic systems to stochastics, Games Theory was on the horizon. And I'm pretty sure Computer Science and Systems Analysis aren't "social" science. :))

I may not have done a good job with the dollar bill example. It was not my intent to claim that an unlikely event which has occured must be random; rather, that once an event has occured, it is no longer a low-probability event. It has, in fact, occured. Hence I argue that an argument based soley (or primarily) upon a "long shot" event is a weak one.

You provide a good example with the oxygen molecules, but what if it occured? The probability is no longer 1:gazillion, but 1.0. Again: my position is that claiming intelligent intervention based purely on the 1:gazillon odds is insufficent. I'd want to see all the data. In this example I would strongly suspect a human intervention, and there would almost certainly evidence thereof. Especially if it were a Columbo episode. :)

And thanks for the heads up; I fixed the /i. Whoops.

Jason: thank you for the clarification. I mis-interpreted your statement. And I still think most philosophers have managed to avoid getting real jobs... ;)

John H: thanks for the kinds words. As I've said before, I have no emotional bias against ID, and I think many people who claim to hold a "scientific" approach to life (which includes many actual scientists) seem to view the method more as a religion, including dogma, prophets, and holy writ.

Actually I think it would be cool to discover compelling objective evidence of God. I just hope -if we did- that he doesn't disappear in a puff of logic! Heh.

Game theory's been applied to studying how likely two countries are to go to war, and in predicting some human behaviors. So yes, it has uses in social sciences. Heck, if you saw "A Beautiful Mind" you saw how the mathematician came up with it by analyzing how to mathematically optimize the odds of he and his friends getting laid. ;-)

If the Intelligent Design theory is not based on faith, but instead another scientific theory, as is claimed, then schools must also allow teachers to teach that the world was created by the Flying Spaghetti Monster, as it is also based on science, not on faith.

A couple of other thoughts:

  • I believe in God. I also think that the theory of evolution is the best way that we have of explaining the world around us as we perceive it. Why does it have to be one or the other?
  • That said, I have no problem teaching ID (or FSM) in any school as long as our children are exposed to it in a class on comparative religions, not in a science classroom.
  • As it is, our nation is going to be hard-pressed to compete with China, India, and the rest without having to starve our childrens' minds with pseudo science.
  • P.S. Anyone who believes that ID is not science, and yet debates proponents of ID, will never win that debate. Ever.

    CT, the problem with the dollar bill example is defining the question. If the question is, "how probable is it that life can exist?" the answer is obviously going to be 1.0. If the question is, "ow probable is it that the diversity of life was established via the mechanisms described by the theory of Evolution?" then that is a much different question.

    Casey Tompkins:

    Phelps,

    I am afraid that you don't grasp the concept behind probabilities. It doesn't matter how you phrase or define the question; any event which has already occured has a probability of 1.0.

    You can posit intelligent design, or you can posit random chance. In either case, the probability is still 1.0. It has already happened.

    Dean's example of the stone which appears to be worked flint is a good one, but that relies on established data. We have flint arrowheads (and other tools) which we know to have been man-made, and we can even reproduce modern examples using known techniques. We also have raftloads of data on how water erosion works on various minerals.

    In other words, we have a large quantity of established data with which we can attempt to match newly found stones, in order to determine whether they were man-made, or not.

    In fact, I can't even imagine how you would manage to estimate your proposed event, since we don't have any data to establish freqencies of occurence. We don't know how various elements of cells developed, how sex was invented, or even how single cells differentiated into specialized cells, and became grouped into a multi-cell animal. We can guess, but it's a tad difficult to experiment. :)
    Dean did recommend William Dembski, whom I have not yet had the chance to read...

    Post a comment

    About

    This page contains a single entry from the blog posted on October 4, 2005 11:26 PM.

    The previous post in this blog was Blue-ray and HD-DVD: wave of the future, or already washed up?.

    The next post in this blog is And now for something Purely Commercial....

    Many more can be found on the main index page or by looking through the archives.

    Powered by
    Movable Type 3.33