hawke123
Registered: 01/29/09
Posts: 669
|
| | 09/16/09 at 11:36 AM | Reply with quote | #1 |
|
I can't be the only person to have noticed this but...
Is it just me or does the scientific method when making predictions commit the logical fallacy of affirming the consequent?
1) If P then Q
2) Q
3) Therefore P
For example, if theory X were true then we would expect to find Y. We do find Y. Therefore, X must be true.
If I am missing something here (which I think I am) I ask that someone explain where my misunderstanding is.
__________________
"Ultimately, the problem with man is not the absence of evidence, it is the suppression of it." - Ravi Zacharias
“Truth is so obscured nowadays and lies [are] so well established that unless we love the truth we shall never recognize it.” - Blaise Pascal |
| Loading... | |
hatsoff
Registered: 08/01/08
Posts: 1,291
|
| | 09/16/09 at 11:42 AM | Reply with quote | #2 |
|
Try thinking of it in terms of Bayesian probability....
In deductive logic, we might say that "if R then Q". But let's instead say
P(Q|R)>P(Q|~R).
Suppose we observe condition Q. Now what can we say about R?
__________________
http://benwallis.blogspot.com |
| Loading... | |
lancia
Registered: 12/08/08
Posts: 744
|
| | 09/16/09 at 01:11 PM | Reply with quote | #3 |
|
Quote: Originally Posted by hawke123 I can't be the only person to have noticed this but...
Is it just me or does the scientific method when making predictions commit the logical fallacy of affirming the consequent?
1) If P then Q
2) Q
3) Therefore P
For example, if theory X were true then we would expect to find Y. We do find Y. Therefore, X must be true.
If I am missing something here (which I think I am) I ask that someone explain where my misunderstanding is.
Yes, I think you are missing something. The scientific method is called the hypothetico-deductive method, and it is based on proving a hypothesis false, not true. In fact, the whole approach is sometimes called falsificationism. So, the inference looks like this. 1) If P, then Q 2) Not Q 3) Therefore not P Proving P true is not possible. The best one can do is to prove it false (but also consider the Duhem-Quine concept about proving something false), on the one hand, or to confirm it to an extent, on the other hand. |
| Loading... | |
hawke123
Registered: 01/29/09
Posts: 669
|
| | 09/16/09 at 01:20 PM | Reply with quote | #4 |
|
But why use such a methodology? Just because you have failed to falsify a theory is not positive evidence that it is true. This would be an argument from ignorance. Why is proving P not possible?
__________________
"Ultimately, the problem with man is not the absence of evidence, it is the suppression of it." - Ravi Zacharias
“Truth is so obscured nowadays and lies [are] so well established that unless we love the truth we shall never recognize it.” - Blaise Pascal |
| Loading... | |
wonderer
Registered: 09/08/08
Posts: 2,835
|
| | 09/16/09 at 01:35 PM | Reply with quote | #5 |
|
Look into antifoundationalism.
In short, to use logic you have to start with something held to be axiomatically true without being able to prove the truth of said axioms with the logic being used.
Given that axioms might be false, an understanding that avoids internal contradiction provides reason for confidence in the axioms even though the axioms must stiil be held open to question.
__________________
“It is in the admission of ignorance and the admission of uncertainty that there is a hope for the continuous motion of human beings in some direction that doesn't get confined, permanently blocked, as it has so many times before in various periods in the history of man." - Richard Feynman |
| Loading... | |
lancia
Registered: 12/08/08
Posts: 744
|
| | 09/16/09 at 01:42 PM | Reply with quote | #6 |
|
Quote: Originally Posted by hawke123
But why use such a methodology? Just because you have failed to falsify a theory is not positive evidence that it is true.
Yes, indeed. But the thought is that by continually rejecting false hypotheses, one comes closer to the true one. Quote: Why is proving P not possible?”
Well, you answered it: one would be committing the fallacy of affirming the consequent. It is not possible because, theoretically anyway, there are many other hypotheses P from which one could deduce the same test implication or prediction Q. So if empirical observation shows that Q is true, then that true observation establishes only that one of the P is true, not necessarily the one being tested. For example, as I have read in one elementary discussion of the fallacy of affirming the consequent, one may say that if there is an eclipse of the sun (P), the streets will be dark (Q). Then upon observing that the streets are in fact dark, one might be tempted to say, committing the fallacy of affirming the consequent, that there is therefore an eclipse of the sun. But that this conclusion is a fallacy is shown by the many other hypotheses for the streets being dark. Maybe it’s very cloudy, maybe there is smoke in the sky, maybe one’s eyes are shaded by dark sunglasses, or maybe it’s simply night-time. |
| Loading... | |
jbejon
Registered: 12/03/07
Posts: 1,027
|
| | 09/16/09 at 05:31 PM | Reply with quote | #7 |
|
I think this question highlights the fact that science works on the basis of probabilities rather than logical inferences.
I haven't personally thought about this issue a great deal. But I think I agree with Hatsoff when he says that if
then R is a decent candidate for an explanation of Q (he didn't say exactly this, but I assume this is what he's driving at). Presumably, then, all other things being equal, the higher the value of (P(Q|R) / P(Q|~R)), the better the explanation?
(I add the ceterus paribus caveat because I think we also have to consider the intrinsic probability of R. To illustrate this fact: Suppose I toss a coin 10 times in a row, and suppose I come up with 8 heads. This is an improbable result, right? So let Q be my tossing 8 heads, and let R be the hypothesis that I've used a loaded coin. P(Q|R), it seems, is greater than P(Q|~R). But it would obviously be wrong to jump to conclusions on this basis. It seems, then, that we should also consider other factors (like the integrity of the coin-tosser, the likelihood of his being able to use a loaded coin, his motivation (or lack of motivation) for doing so, and so on) before concluding that said coin-tosser is using a loaded coin (as well, of course, as mentioned above, as competing hypotheses). |
| Loading... | |
hawke123
Registered: 01/29/09
Posts: 669
|
| | 09/16/09 at 06:07 PM | Reply with quote | #8 |
|
It seems to me that science's methodology is good at falsifying hypotheses but it seems odd to establish something as true because you are unable to falsify it (demonstrate that Q is false).
__________________
"Ultimately, the problem with man is not the absence of evidence, it is the suppression of it." - Ravi Zacharias
“Truth is so obscured nowadays and lies [are] so well established that unless we love the truth we shall never recognize it.” - Blaise Pascal |
| Loading... | |
wonderer
Registered: 09/08/08
Posts: 2,835
|
| | 09/16/09 at 07:12 PM | Reply with quote | #9 |
|
Quote: Originally Posted by hawke123
It seems to me that science's methodology is good at falsifying hypotheses but it seems odd to establish something as true because you are unable to falsify it (demonstrate that Q is false).
It does seems odd to most people to think about it this way until they get used to it.
One important reason for looking at scientific theories in this way is related to the weakness of induction compared to deduction. Scientific claims are necessarily based on induction somewhere along the line. Holding scientific theories as provisional and subject to deductive falsification, acknowledges the inductive nature of scientific theories.
__________________
“It is in the admission of ignorance and the admission of uncertainty that there is a hope for the continuous motion of human beings in some direction that doesn't get confined, permanently blocked, as it has so many times before in various periods in the history of man." - Richard Feynman |
| Loading... | |
bcaserto
Registered: 03/13/08
Posts: 310
|
| | 09/17/09 at 01:36 PM | Reply with quote | #10 |
|
I found this useful: http://falcon.jmu.edu/~omearawm/deduction.html
In medicine both induction and deduction is used. For example:
1. Joe has yellow sclera
2. People with liver failure have yellow sclera
3. Therefore Joe has liver failure
the above is true but is an example of affirming the consequent
the following is also true:
1. Joe has yellow sclera
2. People with hemolytic anemia have yellow sclera
3. Therefore Joe has hemolytic anemia
Both examples are duductively correct, but neither can make a positive diagnosis about Joes condition because both are true. This is a drawn out version of what goes on in a doctors mind when they see a patient. So how do you figure out what is really happening with Joe?
After diagnostic tests are run the results are as follows:
1. Joe has a normal blood cell morphology
2. Joe has normal numbers of blood cells
3. Joe has elevated ALT and AST enzymes in his serum (indicating liver damage)
4. Joe has an antibody titer to Hepatitis C
5. Joe has reduced liver function
Based on 3,4, and 5 we can conclude that Joe probably has liver disease, by induction
Also, by deduction we can conclude that Joe does not have hemolytic anemia based on 1, and 2.
We can also test for the presence to the virus itself to confirm our suspicion that Joe has a Hepatitis C infection. This would be direct evidentiary support the presence of the virus. But this does not necessarily mean that the virus is causing the disease. This is one of the drawback of biology, that there is so much variation and complexity in these systems that establishing cause and effect are troublesome.
By induction though, we can conclude that Joe probably has liver disease due to hepatitis C virus infection because of his clinical symptoms, his bloodwork, and the detection of the virus. we can have even more confidence in this diagnosis because we ruled out hemolytic anemia as a cause of his jaundice.
In many cases we can test our hypothesis by treating for the disease. If the treatment works, it can raise our confidence about our conclusion for future cases.
|
| Loading... | |
lancia
Registered: 12/08/08
Posts: 744
|
| | 09/17/09 at 02:09 PM | Reply with quote | #11 |
|
Quote: Originally Posted by bcaserto
For example:
1. Joe has yellow sclera
2. People with liver failure have yellow sclera
3. Therefore Joe has liver failure
the above is true but is an example of affirming the consequent
the following is also true:
1. Joe has yellow sclera
2. People with hemolytic anemia have yellow sclera
3. Therefore Joe has hemolytic anemia
Both examples are duductively correct
I disagree that these examples are deductively correct. They are deductively incorrect, for if they are examples of a fallacy, the fallacy of affirming the consequent, they can't be both correct and fallacious. |
| Loading... | |
bcaserto
Registered: 03/13/08
Posts: 310
|
| | 09/17/09 at 02:57 PM | Reply with quote | #12 |
|
I see what you mean. I guess they are both fallacious, but they can still be used to set the rest of the illustration, even if they are not examples of sound deductive reasoning.
maybe a better example of deduction would be:
1. Elevated serum ALT indicates hepatocellular injury
2. Joe has elevated serum ALT
3. Therefore Joe has hepatocellular injury
Yes?
|
| Loading... | |
lancia
Registered: 12/08/08
Posts: 744
|
| | 09/17/09 at 03:07 PM | Reply with quote | #13 |
|
Quote: Originally Posted by bcaserto
1. Joe has yellow sclera
2. People with liver failure have yellow sclera
3. Therefore Joe has liver failure
the above is true but is an example of affirming the consequent
This example can be restructured to illustrate the fallacy of affirming the consequent, but since it has no explicit consequent, it is not clear that it is an example of the fallacy of affirming the consequent. It needs an explicit consequent, such as in this example. - If Joe has liver failure, he would have yellow sclera.
- Joe has yellow sclera.
- Therefore, Joe has liver failure.
The conclusion (3) is not deductively valid, but the inference does show clearly the fallacy of affirming the consequent. The clause, “he would have yellow sclera” of premise 1 is the consequent that is affirmed in premise 2. The clause “If Joe has liver failure” of premise 1 is called the antecedent. The inference is incorrect because other conditions (i.e., antecedents) yield the same consequent he would have yellow sclera. |
| Loading... | |
lancia
Registered: 12/08/08
Posts: 744
|
| | 09/17/09 at 03:20 PM | Reply with quote | #14 |
|
Quote: Originally Posted by bcaserto
I see what you mean. I guess they are both fallacious, but they can still be used to set the rest of the illustration, even if they are not examples of sound deductive reasoning.
maybe a better example of deduction would be:
1. Elevated serum ALT indicates hepatocellular injury
2. Joe has elevated serum ALT
3. Therefore Joe has hepatocellular injury
Yes?
You need to provide more information to make this a correct deductive inference. It would be correct if elevated serum ALT indicates ONLY hepatocellular injury. But if elevated serum ALT indicates any other condition, too, then the above would not be valid. |
| Loading... | |
bcaserto
Registered: 03/13/08
Posts: 310
|
| | 09/17/09 at 03:36 PM | Reply with quote | #15 |
|
| It is very specific for hepatocellular injury. Im not sure its necessary to be so explicit. |
| Loading... | |
