Appeal to Authority, for John Powell

[Sheepdog]

since that post was well buried in the thread (i honestly would not have known you answered that post had you not PMed me), so i took your advice, John, and posted here.

quoted from here:

POWELL:
The relevant section is tagged with "SHEEPDOG / JUSTIN." My comments follow.

“
SHEEPDOG / JUSTIN:
Appeal to Authority Fallacy (or, Appeal to False Authority, or Appeal to Questionable Authority)

Let's face it. There is so much information out there, that not one person will ever become qualified as an expert in every field of research relevant to Apologetics. A person can spend decades studying only one field alone. For example, someone can spend half their life studying Bible historicity. Therefore, most laypeople will have to rely on experts from time to time to establish a point (Or, even experts in one field of study will appeal to experts in another field).

POWELL:
This helps to explain why appeals to authority are so important.

ok. no disagreement so far.

SHEEPDOG / JUSTIN:
For this reason, many people become credited scholars in one are a select few areas of study. Then, we can draw from the hard work they have done what we need to defend our arguments. Appealing to an authority to make a point is generally an acceptable way to make a point.

POWELL:
True, but I don't think you understand what is meant by calling something a "logical fallacy."

a logical fallacy is any logical argument that is doesn't produce a true conclusion because of a false or invalid inference. (note, both the terms "false" and "invalid" are important here, as we shall see...)

In fact, Justin, you seem to be in the camp of logicians who are trying to rewrite the definition of "appeal to authority" which tends to unnecessarily confuse the issue. You're trying to equate "appeal to authority" with "appeal to bad authority" and suggesting there's an "appeal to good authority."

While I question the accuracy of your description here, note also that revising our understanding of something based on new insights is not of itself wrong. I'll agree that novelty doesn't make something true either, but keep that in mind.

Tell me, Justin, is it a logical fallacy to appeal to a true authority? If no, then all you really seem to be saying is that some authorities are more reliable than others.

Ok. How about this, why do you trust the word of scientists who debunk the majority of the "evidence" for space aliens visiting earth, over the Alien conspiracy theorists? I have little doubt you'd say that the scientists have become qualified to work in their respective fields, that after scrutinizing the evidence, they would present reliable opinions on the subject. on the other hand, the conspiracy theorists are long on assertions and paranoia, while they are short on evidence.

thus, i have no doubt that you would agree that some authorities are more reliable than others, and their profesional opinions can be weighed as evidence (though not proof, more on this below).

The reason argumentum ad vericundiam, the Latin source for "appeal to authority," was considered a logical fallacy was because one cannot be certain that what a respected person claims to be true is, in fact, true. This situation applies whether the authority is "true" or "false." That's why all appeals to authority are "deductive fallacies." In other words, the following is not a valid deductive argument.

J1. EA is an expert authority in subject S.
J2. EA believes P about subject S.
J3. therefore, P (is true).

First, i'd like to see more on the thought that argumentum ad vericundiam was considered a deductive fallacy (and how you plan on doing this without appealing to an authority yourself, would be interesting to see). but even so, it seems more reasonable to classify Appeal to Authority as inductive reasoning, not deductive reasoning.

first off, you are correct that any appeal in the form of the sylogism above would fallacious. it does not follow that something is true simply because an expert says so. however, how often do you really see an appeal to authority in the form you lay out above? aside from crackpots and such who take word of mouth as god-spoken truth, most appeals i've seen have been similar to form of, "A is more likely to be true than not because E suggests A based on E's study on the subject." Of course, it is the natural tendancy of people for brevity or stating something more extreme than it is, so often it does look like deductive logic.

Second, the sources i checked with classify it as an informal fallacy. but since that's an appeal to authority, let's move on

third, authority by its very nature is not zero-sum. i can be considered an authority on basic engineering concepts. however, don't rely on me to design a bridge-- you go to a professional civil engineer for such a task. likewise, the civil engineer has (or should have) a whole slew of reference material from Ph.D.'s and such which he has to rely on. And no doubt those Ph.D.s got their higher education from somewhere... So, in short, while no authority should be the end all, "it is true cause he says so,"* it is clear that degrees of reliability exists in authority.

*(perhaps unless, God Himself spoke to you directly. but even then, you would need something other than His word of mouth to know what you heard was actually God)

so now, the question that remains is, why does it have to be deductive to be fallacious? the definition i gave above (based on the dictionary.com definition, BTW) would suggest that the phrase "logical fallacy" is not restricted only to formal logic, in that reasoning can be fallacious based on its content, and not just the validity of the logic itself. (though as an aside, not all formal fallacies restricted to deductive logic nore informal to inductive. e.g. question begging is informal yet often plagues deductive reasoning).

SHEEPDOG / JUSTIN:
An Appeal to Authority fallacy, however, occurs when one uses a claim from an person not qualified to comment on the topic one is making the argument in.

POWELL:
But Justin, even those you think are "qualified to comment" might be wrong. Furthermore, who are you to say someone isn't qualified to comment? If the most knowledgeable person available on that subject is P, then maybe you should go ask P, don't you think? Young children are encouraged to rely heavily on what their parents claim to be true regardless of the fact that parents are not people you would consider to be "qualified to comment."

hence you see the inductive nature of taking authoritative opinion as evidence: think of it this way, until you are older, the best source of reliable information you know of is your parents. did you trust your parents when they told you not to cross the road? no doubt you did. why? because they were older, and thus there was good reason to believe they had a good reason for saying so (and it turns out your inductive reasoning was right, as you become aware of how dangerous the road can be for young children, small animals, etc.).

SHEEPDOG / JUSTIN:
For example:

1. Albert Einstein stated that, "God does not play dice."
2. Therefore, random probability (luck) is in contradiction to God's sovereign plan.

This argument is in error because, though Einstein was an expert in science (Heck, he formulated General Relativity Theory), he is not qualified as an expert in theology. Premise (2) may indeed be correct (I have my reservations against commenting either way), but this argument is nevertheless invalid.

”


POWELL:
Who told you that Einstein was not an expert on God? The man was one of the smartest of the last century. If he could figure out relativity, maybe he figured out the mind of God better than you think.

The impression I got from he was not privy to God in the Judeo-Christian or Muslim type of theology. I could be wrong (heh, i vaguely remember something about him speaking positively of the Buddhist view. i dunno). regardless, it is clear that his strong suit was science, not theology. perhaps he did have some insight into God that i am not aware of, but i'd tend demand more be presented from Einstein for his theology than his general relativity theory.

I think you're confused about this issue, Justin. Is the following argument valid?

J4. Albert claims that Albert does not play with dice.
J5. therefore, Albert does not play with dice.

Surely you would think Albert would be an authority on his dice playing. If you think it is a valid argument, then you don't understand what constitutes a "valid deductive argument." You would need to add another premise like "Albert never lies" for it to be "valid."

Or recognize that the argument is inductive, not deductive. once you said "claims," it is clear that your logic is based on observation, not on a law or principle. (to be fair, though, my original example has the appearance of a deductive argument with a missing premise). Arguments which are infered from observation are inductive, while one infered from a law or rule (e.g. Newton's Law of Gravity) is deductive.

SHEEPDOG / JUSTIN:
As a matter of fact, the example above also commits the Quoting Out of Context fallacy, because as I recall, the statement by Einstein was actually not intended as a theological point, but a response to Quantum Theory, which was just beginning to gain scientific acceptance at the time.

POWELL:
I think you are mistaken about Einstein's intent. I think he did mean it to be theological.

hrm. i got the impression from what i read that it was a rather eloquent expression of his opinion of QM. i'll go double check and see if i need to revise that comment...

SHEEPDOG / JUSTIN:
Another kind of appealing to false authority is to state that something is validated by "many" or "most" of a certain kind of authority. For instance, one can say, "the majority of scholars think that the historical Jesus was just a really good teacher." However, unless actual authorities are cited, this appeal is fallacious, for most philosophers say so. (take tongue and plant firmly in cheek).

”


POWELL:
Are you of the opinion, Justin, that if you cite what you and your listener agree are the majority of authorities then your argument becomes valid / nonfallacious?

the last statement was entirely for humor purposes. meh. sometimes my humor is so subtle that even i don't get it

What do you think "valid" and "fallacious" mean?

i believe that has been answered above. note that "fallacious" is not the antinym of "valid": validity has to do with the coherence of the logic. (An argument can be valid, yet have a false premise).

[Sheepdog]

BTW, pardon me while i correct myself on one point. i was technically aware that you did respond to my post. you see, BeHereNow was curious about my real name (Justin), because he thought i might have been someone else he was familiar with. when asked about how he knew my name was Justin, he linked me to Powell's post. so, i did see that Powell responded to me (somewhere), but for some reason it didn't click in my head to actually read through the whole post. sorry about that.

[John Powell]

POWELL:
Let me thank you, Justin, for responding. This is one of my "soap box" issues, so I hope I can make some points with someone as influential as you are in teaching about logical fallacies and with any who might be reading.

I will tag quotes from your web article with "JUSTIN" and comments you made here in TWEB as "SHEEPDOG."

JUSTIN:
For this reason, many people become credited scholars in one are [or?] a select few areas of study. Then, we can draw from the hard work they have done what we need to defend our arguments. Appealing to an authority to make a point is generally an acceptable way to make a point.

POWELL:
True, but I don't think you understand what is meant by calling something a "logical fallacy."

SHEEPDOG:
a logical fallacy is any logical argument that is [sic] doesn't produce a true conclusion because of a false or invalid inference. (note, both the terms "false" and "invalid" are important here, as we shall see...)

POWELL:
That's close. An invalid / fallacious deductive argument is one that could have a true conclusion, but it isn't assured to do so, even if the premises were true. It won't NECESSARILY produce a true conclusion if the premises are true.

Remember, a valid deductive argument is not one that has a true or false conclusion, but one that MUST have a true conclusion, CANNOT have a false conclusion, if the premises are true.

Also, remember that deductive arguments are not "true" or "false" but they are "valid" or "invalid." It is propositions (premises or conclusion) that are "true" or "false."

As I suspected, what you are calling a "logical fallacy" should be more properly called a "deductive fallacy." In other words, it's an argument that would be "invalid," namely one that would not produce a valid deductive argument, that is, one in which the conclusion must be true if the premises were true.

POWELL:
In fact, Justin, you seem to be in the camp of logicians who are trying to rewrite the definition of "appeal to authority" which tends to unnecessarily confuse the issue. You're trying to equate "appeal to authority" with "appeal to bad authority" and suggesting there's an "appeal to good authority."

SHEEPDOG:
While I question the accuracy of your description here, note also that revising our understanding of something based on new insights is not of itself wrong. I'll agree that novelty doesn't make something true either, but keep that in mind.

POWELL:
Good point. However, given the long history of argumentum ad vericundiam, and the general lack of discussion on what's happening behind the scenes, I think as a writer on logic you should make sufficient effort to clarify that you are going along with those logicians who are trying to CHANGE THE MEANING of "appeal to authority."

I wonder how many smart aleck logic students asked their philosophy teachers "If appeal to authority is fallacious, then why should we believe you?" before the teachers decided to adjust things. However, in trying to redefine "appeal to authority" to be "appeal to bad authority" they are obscuring why any appeal to authority was considered fallacious to begin with. They do not produce valid deductive arguments unless you add a false premise like "authorities never lie."

POWELL:
Tell me, Justin, is it a logical fallacy to appeal to a true authority? If no, then all you really seem to be saying is that some authorities are more reliable than others.

SHEEPDOG:
Ok.

POWELL:
Does this mean you think an appeal to true authority is a logical fallacy? What you say below suggests you accept that it's an invalid argument.

SHEEPDOG:
How about this, why do you trust the word of scientists who debunk the majority of the "evidence" for space aliens visiting earth, over the Alien conspiracy theorists? I have little doubt you'd say that the scientists have become qualified to work in their respective fields, that after scrutinizing the evidence, they would present reliable opinions on the subject. on the other hand, the conspiracy theorists are long on assertions and paranoia, while they are short on evidence.

POWELL:
I trust scientists because they have demonstrated sufficiently to me that they are trustworthy. I believe that if I were to follow their methodology I would confirm their results. I rely on them based on an inductive / statistical argument. Of more relevance, this does not mean that such an appeal to authority is a valid deductive argument since even the most reliable authorities can be wrong.

SHEEPDOG:
thus, i have no doubt that you would agree that some authorities are more reliable than others, and their profesional opinions can be weighed as evidence (though not proof, more on this below).

POWELL:
Agreed.

To me "evidence" is any information that helps to form an opinion. I disagree with people who say there is absolutely no evidence for things like aliens, ESP, God, etc. I agree that the evidence is poor, significantly worse than the evidence in support of their critics.

A sound deductive argument (a valid deductive argument with true premises) is an example of a "proof." Because you agree that even an appeal to true authority does not prove the opinion to be true, this suggests you agree that even an appeal to a true authority would be an invalid deductive argument.

POWELL:
The reason argumentum ad vericundiam, the Latin source for "appeal to authority," was considered a logical fallacy was because one cannot be certain that what a respected person claims to be true is, in fact, true. This situation applies whether the authority is "true" or "false." That's why all appeals to authority are "deductive fallacies." In other words, the following is not a valid deductive argument.

J1. EA is an expert authority in subject S.
J2. EA believes P about subject S.
J3. therefore, P (is true).

SHEEPDOG:
First, i'd like to see more on the thought that argumentum ad vericundiam was considered a deductive fallacy (and how you plan on doing this without appealing to an authority yourself, would be interesting to see).

POWELL:
GOOD POINT! You see an important problem with claiming appeals to authority are "bad" reasoning. Again, I support appeals to authority based on statistical reliability. I concede that they do not produce valid deductive arguments (without adding a false premise.)

I'm not sure of the history of logic, but I've been led to believe that early logicians concentrated on arguments with certain conclusions. I don't know when these were called "deductive," but perhaps it was when they were contrasted with "inductive" arguments. "Deductive" arguments were making inferences of particulars (such as whether Socrates was mortal) based on general principles (all men are mortal). Inductive arguments worked the reverse way. You come up with general principles, laws of nature, based on individual observations.

Scientists effectively slapped logicians in the face by demonstrating the importance of inductive reasoning to increase our knowledge of the universe. Too strict reliance on deductive forms of reasoning had misled philosophers to believe that the heavier of two iron balls would hit the ground first. A simple demonstration by Galileo proved an important weakness of logic sans experiment.

Because this happened many more times, non scientist philosophers have been forced, reluctantly apparently, to include induction as a kind of "good" reasoning, but too many of them still tend to think of it as inferior to deductive reasoning. Science uses both kinds.

Teachers of deductive logic categorized common arguments which did not produce certain conclusions as "fallacies," such as argumentum ad vericundiam, ad misericordiam, ad baculum, ad populam, etc.

The argumentum ad vericundiam or "appeal to awe" was the idea that just because someone awesome (such as the king) believed something did not necessarily mean it was true. This is what is meant by an appeal to authority.

SHEEPDOG:
but even so, it seems more reasonable to classify Appeal to Authority as inductive reasoning, not deductive reasoning.

POWELL:
Now we're talking! Not enough debaters realize the difference. They have been misled by reading fallacy lists such as your own to remain ignorant of the value of inductive / statistical arguments. When someone uses one of the fallacies, their critic often points it out even if the argument should have been a statistical one which would make the charge of fallacy to be inappropriate. Because the proponent may be similarly ignorant, he is unable to properly defend his own argument.

There is a problem for you, however, Justin. When you call appeal to authority an inductive / statistical argument then it is neither "fallacious / invalid" nor "valid," but "weak" or "strong." Appealing to a true authority would typically be a strong inductive argument (greater than 50% probability of having a true conclusion) whereas an appeal to a false authority would typically be a weak inductive argument (less than 50% probability of having a true conclusion).

SHEEPDOG:
first off, you are correct that any appeal in the form of the sylogism above would fallacious. it does not follow that something is true simply because an expert says so. however, how often do you really see an appeal to authority in the form you lay out above? aside from crackpots and such who take word of mouth as god-spoken truth, most appeals i've seen have been similar to form of, "A is more likely to be true than not because E suggests A based on E's study on the subject." Of course, it is the natural tendancy of people for brevity or stating something more extreme than it is, so often it does look like deductive logic.

POWELL:
I agree that using syllogisms is uncommon. However, it's my impression that arguments in discussion forums DON'T use statistical terms (like "probably") often enough. The form is generally "P says Q." with the implication that "Q (is true)."

SHEEPDOG:
Second, the sources i checked with classify it as an informal fallacy.

POWELL:
I think that means it's not a formal fallacy. A "formal" fallacy is one where it's the form of the argument that is invalid such as "affirming the consequent." In those cases it doesn't matter what "p" and "q" stand for, the argument cannot be a valid one. An informal fallacy is one in which it's not the form that causes it to be invalid, but I guess the substance.

SHEEPDOG:
but since that's an appeal to authority, let's move on

POWELL:
I'm happy that you are realizing how often one appeals to authority. Too many people don't realize that.

SHEEPDOG:
third, authority by its very nature is not zero-sum. i can be considered an authority on basic engineering concepts. however, don't rely on me to design a bridge-- you go to a professional civil engineer for such a task. likewise, the civil engineer has (or should have) a whole slew of reference material from Ph.D.'s and such which he has to rely on. And no doubt those Ph.D.s got their higher education from somewhere... So, in short, while no authority should be the end all, "it is true cause he says so,"* it is clear that degrees of reliability exists in authority.

POWELL:
EXACTLY! No one is an absolute authority on an issue and those in distantly related fields or those who have read about it might have "authoritative" knowledge to share.

SHEEPDOG:
*(perhaps unless, God Himself spoke to you directly. but even then, you would need something other than His word of mouth to know what you heard was actually God) so now, the question that remains is, why does it have to be deductive to be fallacious? the definition i gave above (based on the dictionary.com definition, BTW) would suggest that the phrase "logical fallacy" is not restricted only to formal logic, in that reasoning can be fallacious based on its content, and not just the validity of the logic itself. (though as an aside, not all formal fallacies restricted to deductive logic nore informal to inductive. e.g. question begging is informal yet often plagues deductive reasoning).

POWELL:
Some logicians want to use the more general meaning of "fallacious" as "bad," but I don't suggest that. Too many people use it to mean "invalid deductive argument" even if they don't realize that.

I posted previously in the philosophy section of TWEB my claim that Modus Ponens arguments when claimed to be valid (rather than sound) are essentially circular arguments. The question begging is built into the form.

JUSTIN:
An Appeal to Authority fallacy, however, occurs when one uses a claim from an [a] person not qualified to comment on the topic one is making the argument in.

POWELL:
But Justin, even those you think are "qualified to comment" might be wrong. Furthermore, who are you to say someone isn't qualified to comment? If the most knowledgeable person available on that subject is P, then maybe you should go ask P, don't you think? Young children are encouraged to rely heavily on what their parents claim to be true regardless of the fact that parents are not people you would consider to be "qualified to comment."

SHEEPDOG:
hence you see the inductive nature of taking authoritative opinion as evidence: think of it this way, until you are older, the best source of reliable information you know of is your parents. did you trust your parents when they told you not to cross the road? no doubt you did. why? because they were older, and thus there was good reason to believe they had a good reason for saying so (and it turns out your inductive reasoning was right, as you become aware of how dangerous the road can be for young children, small animals, etc.).

POWELL:
Exactly. It's a statistical argument, not a sure thing.

JUSTIN:
For example:

1. Albert Einstein stated that, "God does not play dice."
2. Therefore, random probability (luck) is in contradiction to God's sovereign plan.

This argument is in error because, though Einstein was an expert in science (Heck, he formulated General Relativity Theory), he is not qualified as an expert in theology. Premise (2) may indeed be correct (I have my reservations against commenting either way), but this argument is nevertheless invalid.

POWELL:
Who told you that Einstein was not an expert on God? The man was one of the smartest of the last century. If he could figure out
relativity, maybe he figured out the mind of God better than you
think.

SHEEPDOG:
The impression I got from he was not privy to God in the Judeo-Christian or Muslim type of theology. I could be wrong (heh, i
vaguely remember something about him speaking positively of the Buddhist view. i dunno). regardless, it is clear that his strong suit was science, not theology. perhaps he did have some insight into God that i am not aware of, but i'd tend demand more be presented from Einstein for his theology than his general relativity theory.

POWELL:
Right. My point is that when we choose our authorities we don't necessarily ask if they obtained their college degree in that speciality. Smart people tend to be wise about a lot of things unrelated to their primary expertise.

POWELL:
I think you're confused about this issue, Justin. Is the following argument valid?

J4. Albert claims that Albert does not play with dice.
J5. therefore, Albert does not play with dice.

Surely you would think Albert would be an authority on his dice playing. If you think it is a valid argument, then you don't understand what constitutes a "valid deductive argument." You would need to add another premise like "Albert never lies" for it to be "valid."

SHEEPDOG:
Or recognize that the argument is inductive, not deductive. once you said "claims," it is clear that your logic is based on observation, not on a law or principle. (to be fair, though, my original example has the appearance of a deductive argument with a missing premise).

Arguments which are infered from observation are inductive, while one infered from a law or rule (e.g. Newton's Law of Gravity) is deductive.

POWELL:
I think you understand a lot better than I thought.

Critics of arguments should be aware that the argument presented may not have been claimed to be deductive.

Unfortunately, now we have a problem between differing definitions of "deductive" and "inductive." I think "deductive" should mean making inferences on individual cases based on general principles WHETHER THE CONCLUSION IS CERTAIN OR NOT. I think inductive should be understood to be the reverse. It's making inferences about general principles based on individual observations.

However, in beginning logic texts, this is not how "deductive" and "inductive" are defined. I've been reading Copi and Cohen since they were recommended to me by a teacher of philosophy at TOPICA'S II Errancy (Doug Krueger) and it happens to be the text used at my college.

To Copi and Cohen, a deductive argument is one claimed to have a certain conclusion if the premises are true. If the claim of certainty is confirmed then it's a valid deductive argument. If the claim of certainty is not confirmed then it's an invalid deductive argument. Whether these arguments are "general to specific" or "specific to general" is apparently irrelevant.

To Copi and Cohen, an inductive argument is one that is NOT claimed to have a certain conclusion if the premises are true. Again, apparently the general / specific aspect is irrelevant.

I think ALL arguments should be considered what Copi and Cohen call inductive arguments, but called "statistical" instead. I have concluded that even so-called valid deductive arguments are merely statistical in nature. The question is how probable, how close to absolute certainty a conclusion is given true premises, since none, or very few, of them reach absolute certainty.

I posted these arguments at TWEB many months ago.

JUSTIN:
As a matter of fact, the example above also commits the Quoting Out of Context fallacy, because as I recall, the statement by Einstein was actually not intended as a theological point, but a response to Quantum Theory, which was just beginning to gain scientific acceptance at the time.

POWELL:
I think you are mistaken about Einstein's intent. I think he did mean it to be theological.

SHEEPDOG:
hrm. i got the impression from what i read that it was a rather eloquent expression of his opinion of QM. i'll go double check and see if i need to revise that comment...

POWELL:
Please let me know if I was mistaken about Einstein's probable intent.

JUSTIN:
Another kind of appealing to false authority is to state that something is validated by "many" or "most" of a certain kind of authority. For instance, one can say, "the majority of scholars think that the historical Jesus was just a really good teacher." However, unless actual authorities are cited, this appeal is fallacious, for most philosophers say so. (take tongue and plant firmly in cheek).

POWELL:
Are you of the opinion, Justin, that if you cite what you and your listener agree are the majority of authorities then your argument becomes valid / nonfallacious?

SHEEPDOG:
the last statement was entirely for humor purposes. meh. sometimes my humor is so subtle that even i don't get it

POWELL:
I see.

POWELL:
What do you think "valid" and "fallacious" mean?

SHEEPDOG:
i believe that has been answered above. note that "fallacious" is not the antinym of "valid": validity has to do with the coherence of the logic. (An argument can be valid, yet have a false premise).

POWELL:
I think "fallacy" lists are basically using the "fallacious = invalid" definition.

You're right that a valid deductive argument can have a false premise. Were you also aware that a valid deductive argument can have a false conclusion?

A valid deductive argument may have false premises and false conclusion, false premises and true conclusion, or true premises and true conclusion. What it cannot have is true premises and false conclusion.

I'll give examples of each if you need me to.

A valid deductive argument (according to people like Copi and Cohen) is one in which the conclusion must be true, cannot be false, if the premises are true.

John Powell

[Warcraft3]

John Powell and Sheepdog:

Excellent exchange from what Ive read (I did not carefully read all of it due to time constraints)!!

I thought the "appeal to authority" matter was rather obvious. Of course we all do that, it is just a part of life...so I never did see that as a true fallacy.

In fact there are several so-called fallacies that I do not think are as fallacious as is often claimed.

These would include:

1. Personal incredulity
2. Argument from silence
3. Appeal to authority


There are others but I am tired and will continue at a later time.


Russ

[John Powell]

STEADELE:
Excellent exchange from what Ive read (I did not carefully read all of it due to time constraints)!!

I thought the "appeal to authority" matter was rather obvious. Of course we all do that, it is just a part of life...so I never did see that as a true fallacy.

POWELL:
I don't think you quite understand, Russ, and, unfortunately, it's not as obvious as you seem to think.

The appeal to authority (whether good or bad authority) is a deductive fallacy because it is not absolutely certain that what an authority believes to be true is, in fact, true. So, an appeal to authority is a "bad" deductive argument, what's called an "invalid deductive argument" or "invalid argument" for short. For the same reason, those other items listed in fallacy lists are "bad" deductive arguments.

On the other hand, the appeal to authority is one of the best reasons to believe because it can be a "strong" inductive / statistical argument. Most people call them "inductive," but I prefer the term "statistical."

Some logicians may be frustrated by these two facts:

1. Appeal to Authority is a deductive fallacy, so it's "bad."
2. Appeal to Authority can be a strong inductive / statistical argument, so it can be "good," often better than available deductive arguments.

If Justin doesn't see it coming already, he will soon learn that I believe that many of the other "fallacies" in lists like his own are in a similar situation. They may be bad / invalid / fallacious deductive arguments, but they can be good / strong inductive / statistical arguments.

STEADELE:
In fact there are several so-called fallacies that I do not think are as fallacious as is often claimed.

POWELL:
We think a lot alike, Russ.

However, what you mean by "fallacy" as being "bad reasoning" in general is not what those making typical fallacy lists mean by the term (even if they don't realize this). They use the term fallacy to be an invalid deductive argument. A "deductive argument" is one in which it is claimed that the conclusion is absolutely certain of being true if the premises are true. A "valid" deductive argument is one in which the claim of certainty is correct. An "invalid" deductive argument is one in which the claim of certainty is incorrect.

Also, there are not supposed to be "degrees of validity." Either the deductive argument is fully valid / nonfallacious or it is fully invalid / fallacious. It's when you use the term "fallacious" in the more general sense of "bad" that there can be degrees. For example, some statistical arguments are stronger than others (there is a higher probability that the conclusion is true if the premises are true).

Actually, I think there are degrees of validity and that, in fact, even so-called valid deductive arguments are really just statistical arguments with very high probabilities of the conclusion being true if the premises are true. I guess that kind of revolutionary thinking can happen when your background is in science rather than philosophy.

STEADELE:
These would include:

1. Personal incredulity
2. Argument from silence
3. Appeal to authority

POWELL:
A way to test your claim is the next time a critic points out that you're committing a fallacy such as "appeal to authority" or whatever, ask them on what basis they were assuming that you were claiming to be making a valid deductive argument? If all you were doing was making a statistical argument then the charge of "fallacy / invalid" is inappropriate since that only applies to deductive arguments. For statistical arguments, one needs to determine things like whether the statistical argument is strong (the conclusion is more than 50% probable of being true if the premises are true) or weak (less than 50% probability). Surely they realize that you know that an appeal to authority or whatever is an "invalid argument" since your conclusion is not absolutely certain of being true if the premises are true, but who said you claimed it was?

I think you'll find that most of your critics will be poorly equipped to make a proper reply. Some may refuse to argue because you're too rebellious to accept the appeal to authority of fallacy lists. Hypocritical, huh? Others may try to defend absolutest logic against probablistic logic. Then it may become a fight between philosophy and science. Science usually comes out the victor in those debates.

It's too bad, but it's not a requirement that statistical arguments explicitly include probabilistic terms in the conclusion (such as "it is likely the case that") any more than it is required that deductive arguments explicitly include absolutest terms (like "it is necessarily the case that"). I think both kinds of arguments should include these terms somewhere appropriate so it's clear to the reader whether the proponent is making a deductive argument or an inductive / statistical argument.

STEADELE:
There are others but I am tired and will continue at a later time.

Russ

POWELL:
Keep thinking, Russ!

John Powell

[Sheepdog]

hiya John,

i owe you a response, but my schedule is getting rough, and will be even rougher as school starts next week. i should get a reply in either tonight or tomorrow. beyond that, i can't garantee i'll be on TWeb enough to continue this discussion.

looking forward to responding though... i can tell why Holding has you in high esteem.

[John Powell]

POWELL:
Whenever you get the chance. These issues tend to remain interesting weeks and months after they are begun.

I hope I don't do anything to destroy that positive feeling.

John Powell

[Sheepdog]

Hiya John,

Rather than responding point-by-point, i opted for a general response, as it will reduce redundancy and i think we actually have quite a bit of agreement. first off, on your comments regarding my definition of a logic fallacy, you are right on. you articulated the general idea i was going after, perhaps i got a bit loose in my jargon or something. a conclusion may be true, even if the logic is invalid or contains a false premise, but such a conclusion does not follow from a particular argument if it is unsound. Also, i am aware that a deductive argument can be valid, yet unsound, if a premise is false (heh, can you tell i just had a class today on logic?)

After some consideration, I think I will revise the section on Appeal to Authority in my logical fallacies page. The primary change would be to emphasize that no authority is hard proof in the sense that you can make a deductive argument from it. And, if I can get a hold of some documentation, I'll definitely note that Appeal to Authority as defined as a fallacy originated based on the "A said p, therefore p" style of deductive argument. It's not so much i don't trust you John, i just tend to be strict about citing things that aren't general knowledge.

more than likely i'll also review the rest of the "fallacies" on my page to make sure the same sort of ambiguity between deductive and inductive reasoning hasn't been commited elsewhere, and revise as necessary. i was considering writing a page on logic and critical thinking for CARM in the near future, and now that would be more likely.

now one section of you post i'd like to respond to directly:

Powell:
There is a problem for you, however, Justin. When you call appeal to authority an inductive / statistical argument then it is neither "fallacious / invalid" nor "valid," but "weak" or "strong." Appealing to a true authority would typically be a strong inductive argument (greater than 50% probability of having a true conclusion) whereas an appeal to a false authority would typically be a weak inductive argument (less than 50% probability of having a true conclusion).

Here is where we part directions. The two sites I checked (nikzor.org/features/fallacies, gncurtis.home.texas.net), give a more general definition of "fallacy," such that inductively "weak" arguments can be considered fallacious. The dictionary.com (American Heritage Dictionary) definition, which i based mine definition on, reads, "A statement or an argument based on a false or invalid inference." The Webster's gives a definition more along the lines of the sites I mentioned. Now, I would argue, based on this (and keeping with my def., even), that a weak inductive arguement produces a "false" inference, in the sense that the evidence at hand is too weak to support the stated conslusion. This may be where the confusion between us is resulting.

You concede that many logicians do this at one point. Then perhaps what I should do is preamble my page with a discusion of the general use of the term "fallacy" and on deductive vs. inductive reasoning, so that there will be less confusion? At least until i do write up a page on reasoning, if it does come together.

Coincidentally, I just got Copi and Cohen today for a class this semester. So, i'll be using that as a reference as well. Regarding whether there are deductive arguments in existance, I prefer not to comment in full, though a strong case can be made for them from definitions and categories (e.g. all dogs are mammals).

Still nothing on the Einstein thing. i'll probably drop the comment when i revise the section, should i not find the context. Well, i have some xml/xslt pages to write, so peace out for now.

[John Powell]

SHEEPDOG:
Hiya John,

Rather than responding point-by-point, i opted for a general response, as it will reduce redundancy and i think we actually have quite a bit of agreement. first off, on your comments regarding my definition of a logic fallacy, you are right on. you articulated the general idea i was going after, perhaps i got a bit loose in my jargon or something. a conclusion may be true, even if the logic is invalid or contains a false premise, but such a conclusion does not follow from a particular argument if it is unsound.

POWELL:
Instead of "unsound" you probably should put "invalid."

An unsound argument is either a nondeductive argument or a deductive argument that is invalid or a deductive argument that is valid but with false premises.

SHEEPDOG:
Also, i am aware that a deductive argument can be valid, yet unsound, if a premise is false (heh, can you tell i just had a class today on logic?)

POWELL:
Correct.

SHEEPDOG:
After some consideration, I think I will revise the section on Appeal to Authority in my logical fallacies page. The primary change would be to emphasize that no authority is hard proof in the sense that you can make a deductive argument from it.

POWELL:
Instead of "deductive argument" I think you should say "sound deductive argument". You can make valid deductive arguments using "fallacies" but only by introducing premises that are unlikely to be true.

For example.

1. Authorities are always right.
2. Authority A says B.
3. Therefore, B

This is a valid deductive argument (the conclusion must be true, cannot be false if the premises are true), however it is surely unsound because premise 1 is surely false.

SHEEPDOG:
And, if I can get a hold of some documentation, I'll definitely note that Appeal to Authority as defined as a fallacy originated based on the "A said p, therefore p" style of deductive argument. It's not so much i don't trust you John, i just tend to be strict about citing things that aren't general knowledge.

POWELL:
Please don't merely trust me on something as important as this. However, it's my understanding that it went something like this

1. Great person A (e.g., the King) believes P
2. Therefore, P

The argumentum ad vericundiam was the "appeal to awe." Because the person was awesome, they should be believed.

I don't know when this kind of argument was first considered to be deductive.

SHEEPDOG:
more than likely i'll also review the rest of the "fallacies" on my page to make sure the same sort of ambiguity between deductive and inductive reasoning hasn't been commited elsewhere, and revise as necessary.

POWELL:
I would emphasize that these are "deductive" fallacies because one cannot form sound deductive arguments from them. One cannot be absolutely certain that the conclusions based upon them are true.

You might also mention that sometimes some of the deductive fallacies can be strong inductive arguments, justified for making decisions.

SHEEPDOG:
i was considering writing a page on logic and critical thinking for CARM in the near future, and now that would be more likely.

POWELL:
Go for it. If you would like me to proofread your paper, I'm willing. Of course you don't have to do anything I might suggest.

:SHEEPDOG:
now one section of you post i'd like to respond to directly:

Powell:
There is a problem for you, however, Justin. When you call appeal to authority an inductive / statistical argument then it is neither "fallacious / invalid" nor "valid," but "weak" or "strong." Appealing to a true authority would typically be a strong inductive argument (greater than 50% probability of having a true conclusion) whereas an appeal to a false authority would typically be a weak inductive argument (less than 50% probability of having a true conclusion). ”

SHEEPDOG:
Here is where we part directions. The two sites I checked (nikzor.org/features/fallacies, gncurtis.home.texas.net), give a more general definition of "fallacy," such that inductively "weak" arguments can be considered fallacious. The dictionary.com (American Heritage Dictionary) definition, which i based mine definition on, reads, "A statement or an argument based on a false or invalid inference." The Webster's gives a definition more along the lines of the sites I mentioned.

POWELL:
An "invalid inference" refers to deductive arguments. A "false" inference probably means the same thing. You could think of nondeductive arguments as fallacious because none of them can be sound deductive arguments, but that would be like criticizing a woman for being a bad father. An inductive / statistical argument doesn't claim or seek to be valid, so the failure is not too notable. To say that a certain nondeductive argument is fallacious would be redundant, because by the "fallacious = invalid" definition, all nondeductive arguments are fallacious.

If you are using the more general "fallacious = bad" definition then you should call strong inductive / statistical arguments as nonfallacious. These would include many appeals to authority.

SHEEPDOG:
Now, I would argue, based on this (and keeping with my def., even), that a weak inductive arguement produces a "false" inference, in the sense that the evidence at hand is too weak to support the stated conslusion. This may be where the confusion between us is resulting.

POWELL:
Well, you could justifiably say that the weak inductive argument supports the stated conclusion, but does so weakly. It gives some support, right? However, the opposite conclusion is strongly supported, more likely true.

SHEEPDOG:
You concede that many logicians do this at one point. Then perhaps what I should do is preamble my page with a discusion of the general use of the term "fallacy" and on deductive vs. inductive reasoning, so that there will be less confusion?

POWELL:
Yes and occasionally make relevant comments to catch those people who don't read the introductory articles when they surf for information on fallacies.

SHEEPDOG:
At least until i do write up a page on reasoning, if it does come together.

Coincidentally, I just got Copi and Cohen today for a class this semester. So, i'll be using that as a reference as well.

POWELL:
They don't consider all appeals to authority to be fallacious, but they admit that they do not produce a sound deductive argument. They use the more general meaning of "fallacious," but then I'm not sure they're consistent in their fallacy section.

SHEEPDOG:
Regarding whether there are deductive arguments in existance, I prefer not to comment in full, though a strong case can be made for them from definitions and categories (e.g. all dogs are mammals).

POWELL:
Perhaps you'd like to take me up on that later on.

SHEEPDOG:
Still nothing on the Einstein thing. i'll probably drop the comment when i revise the section, should i not find the context. Well, i have some xml/xslt pages to write, so peace out for now.

POWELL:
It's ok to question the correctness of Einstein's theological comments since his expertise was in physics.

John Powell

[Sheepdog]

too late to respond tonight, but one irrelevancy:

Sheepdog:
i was considering writing a page on logic and critical thinking for CARM in the near future, and now that would be more likely.

I meant Tekton (where the fallacies page is at). why i said CARM, i do not know.