Seems like if you apply a few degrees of Kevin Bacon to it I bet you would find that the people doing the shooting
are not all that unrelated with the people doing the slandering and intimidating.
Doubtless. On the other hand, I can also guarantee that the vast majority of persons in the US can be "linked" to said slanderers (in the sense of there being a short path from the one to the other) for the appropriate choice of social relation. This is one problem with the "guilt-by-association" argument: with a distance of only a few hops, virtually anyone can be tied to a very large number of persons. Since most people seem to underestimate the extent to which this is true, such connections can seem like a very impressive indication of guilt (or, at least, suspicious behavior) to the uninitiated.
Of course, that doesn't mean that there isn't a closer connection between those posting doctors' names and addresses on web sites and those gunning them down in cold blood. Just be careful in using relationship chains per se as evidence of complicity*....
-Carter
* And, if you're actually participating in a conspiracy, make sure your position has a low betweenness score.;-) (IIRC, based on some work by Baker and Faulkner some years back.)
Perhaps the saddest thing about all this, IMHO...
on
eFront From Inside
·
· Score: 1
...is what it reveals about the eFront members'
abuse of the English language. At the very least,
one would think that they could conspire in complete sentences.
Sleazy W3 ad barons, runaway IP law, blink tags...this place just hasn't been the same since the Eternal September.
The "philosophical" answer to your question derives (in this case) from the philosophy of inference.
The intelligent design argument, in the form you are invoking, is premised on the following reasoning:
1. We observe our own existence. (Call this "O")
2. Under the hypothesis that the universe is designed for us ("Hd"), it is almost certain that we would observe our own existence. Thus, p(O|Hd)->1. (By this notation, I mean that the probability of O given Hd is arbitrarily close to 1.)
3. Under the hypothesis that the universe is not designed for us, it is almost certain that we would not exist, and thus that we could not observer our own existance. Thus, they argue, p(O|not(Hd))->0
4. The prior probability of design is not arbitrarily small.
There are several potential problems with the above reasoning (for instance, each of the presumptive probabilities can be challenged on a variety of grounds), but it contains at least one fundamental flaw which renders it inapplicable.
(Can you guess where it is?)
(No peeking!)
Ok, here's the deal: the fundamental flaw in question is contained in statement (3), which confuses the probability of our existence with the probability of our observing our existance. These are not synonymous, and indeed this sort of confusion can lead (as in the present case) to all manner of mistaken inferences. In point of fact, p(O|not(Hd)) must also be equal to 1, because our observation of our existence is conditioned on our existence in the first place. If we weren't here, we wouldn't be here to see it.
The effect of this revelation on the above reasoning can be easily demonstrated using Bayes' theorem:
Thus, our self observation tells us nothing, one way or the other about the hypothesis that the universe was designed for us. (Sorry, Charlie Brown.) This observation is a necessity, and carries no information regarding the proposition in question.
As I'm sure others have or will point out, the above argumentation is not in any way new (though it is only rarely presented formally). The phenomenon is a special case of the practice often called "sampling on the dependent variable," and is well understood by competent methodologists. A variant of the phenomenon (sometimes called the "stock market" or "perfect prediction" swindle) has been used to scam consumers; the late Morris DeGroot has a nice description of it in his introductory prob/stat book, but you can probably find other accounts elsewhere if you look for them. An essentially similar logic to used here in refuting the ID argument has been invoked as the "weak anthropic principle" by Carter, Barrow, Tipler, and others, though they later go on to make a number of other rather strange claims under the "anthropic principle" heading which derive little or no justification from the censored sampling phenomenon. You can find these arguments in the books of the above authors, though again one should be careful to seperate their cogent refutation of the argument for ID from their other, far more extreme, claims.
Slashdot Mirror
Doubtless. On the other hand, I can also guarantee that the vast majority of persons in the US can be "linked" to said slanderers (in the sense of there being a short path from the one to the other) for the appropriate choice of social relation. This is one problem with the "guilt-by-association" argument: with a distance of only a few hops, virtually anyone can be tied to a very large number of persons. Since most people seem to underestimate the extent to which this is true, such connections can seem like a very impressive indication of guilt (or, at least, suspicious behavior) to the uninitiated.
Of course, that doesn't mean that there isn't a closer connection between those posting doctors' names and addresses on web sites and those gunning them down in cold blood. Just be careful in using relationship chains per se as evidence of complicity*....
-Carter
* And, if you're actually participating in a conspiracy, make sure your position has a low betweenness score. ;-) (IIRC, based on some work by Baker and Faulkner some years back.)
Sleazy W3 ad barons, runaway IP law, blink tags...this place just hasn't been the same since the Eternal September.
-Carter
The intelligent design argument, in the form you are invoking, is premised on the following reasoning:
1. We observe our own existence. (Call this "O")
2. Under the hypothesis that the universe is designed for us ("Hd"), it is almost certain that we would observe our own existence. Thus, p(O|Hd)->1. (By this notation, I mean that the probability of O given Hd is arbitrarily close to 1.)
3. Under the hypothesis that the universe is not designed for us, it is almost certain that we would not exist, and thus that we could not observer our own existance. Thus, they argue, p(O|not(Hd))->0
4. The prior probability of design is not arbitrarily small.
5. By (1)-(4), together with Bayes' Theorem,
O |not(Hd)))
p(Hd|O)=(p(Hd)p(O|Hd))/(p(Hd)p(O|Hd)+p(not(Hd))p(
--> p(Hd|O) approx (1p(Hd))/(1p(Hd)+0p(not(Hd)))
=1. QED.
There are several potential problems with the above reasoning (for instance, each of the presumptive probabilities can be challenged on a variety of grounds), but it contains at least one fundamental flaw which renders it inapplicable.
(Can you guess where it is?)
(No peeking!)
Ok, here's the deal: the fundamental flaw in question is contained in statement (3), which confuses the probability of our existence with the probability of our observing our existance. These are not synonymous, and indeed this sort of confusion can lead (as in the present case) to all manner of mistaken inferences. In point of fact, p(O|not(Hd)) must also be equal to 1, because our observation of our existence is conditioned on our existence in the first place. If we weren't here, we wouldn't be here to see it.
The effect of this revelation on the above reasoning can be easily demonstrated using Bayes' theorem:
O |not(Hd)))
p(Hd|O)=(p(Hd)p(O|Hd))/(p(Hd)p(O|Hd)+p(not(Hd))p(
--> p(Hd|O) approx (1p(Hd))/(1p(Hd)+1p(not(Hd)))
=p(Hd). QED.
Thus, our self observation tells us nothing, one way or the other about the hypothesis that the universe was designed for us. (Sorry, Charlie Brown.) This observation is a necessity, and carries no information regarding the proposition in question.
As I'm sure others have or will point out, the above argumentation is not in any way new (though it is only rarely presented formally). The phenomenon is a special case of the practice often called "sampling on the dependent variable," and is well understood by competent methodologists. A variant of the phenomenon (sometimes called the "stock market" or "perfect prediction" swindle) has been used to scam consumers; the late Morris DeGroot has a nice description of it in his introductory prob/stat book, but you can probably find other accounts elsewhere if you look for them. An essentially similar logic to used here in refuting the ID argument has been invoked as the "weak anthropic principle" by Carter, Barrow, Tipler, and others, though they later go on to make a number of other rather strange claims under the "anthropic principle" heading which derive little or no justification from the censored sampling phenomenon. You can find these arguments in the books of the above authors, though again one should be careful to seperate their cogent refutation of the argument for ID from their other, far more extreme, claims.
-Carter