The logical fallacies thread

[jBirch]

Originally Posted by Neil

I don't see the point of continuously quoting these fallacies. It's obvious (if you think about it for a second) that it's wrong to extrapolate the behaviour of one person to an entire group that they happen to be part of; you don't need to learn that it's an example of the *whatever* fallacy. We don't need unnecessary jargon for every commonly-encountered piece of dodgy reasoning.

Sure we do! How else are going to train our minds to recognise "dodgy reasoning"? Remember, common sense isn't.

Another thing that annoys me is that people often use these fallacies as a type of point-scoring. Instead of engaging in constructive argument with the other person, they point out a fallacy and then act all smug - on a par with criticising their grammar/typos.

It goes to ethos in an argument. The formal purpose of an argument is to persuade everyone that your opinion is correct. You do this by first setting out your own credentials and making a case for why you need to be listened to. This is called Ethos and gives any opinion on the subject at hand almost the weight of fact. Think Expert Testimony in a court of law. This makes it so that people will listen attentively to what you have to say. You then engage the emotions of your audience and get them to suspend their critical judgement so that your ability to launch a new opinion has a receptive audience. This is called Pathos and makes your audience *want* to accept your opinion without necessarily evaluating it too closely. Finally, you logically tie together the emotion and your credibility into the rationale for why your opinion is right. This is called logos and makes it easy for the audience to satisfy their Pathos. They see the lines from the fact to the opinion and accept them as valid. The speech between Mark Antony and Brutus demonstrates this principal admirably.

On to Inq's specific rhetorical tactic...

He's basically just attacking your ethos and poking holes in your logos. If you engage in irrational behaviour then he'll succeed in persuading the audience that your ethos is shaky and your logos unsupportable. Once he's shown that you are not to be listened to or if you are, then your conclusions are wrong, he starts his own argument. In a rational debate, to counter this tactic you simply refute each of his points persuasively and leave him in a weak position. This strikes at his Ethos and builds up your own. What you have to watch out for is that he doesn't get you so far down the path of burning his straw men that you can't find your way back to your original argument. He's a sly devil, that's for sure...but he adds a lot of quality to our debates here and weeds out the irrational arguments pretty quickly.

That, and he's kinda balding so we can't hurt his feelings too badly...*grin*

James.

[Dr. Pfleschbach]

Originally Posted by jBirch

On to Inq's specific rhetorical tactic...

He's basically just attacking your ethos and poking holes in your logos. If you engage in irrational behaviour then he'll succeed in persuading the audience that your ethos is shaky and your logos unsupportable. Once he's shown that you are not to be listened to or if you are, then your conclusions are wrong, he starts his own argument. In a rational debate, to counter this tactic you simply refute each of his points persuasively and leave him in a weak position. This strikes at his Ethos and builds up your own. What you have to watch out for is that he doesn't get you so far down the path of burning his straw men that you can't find your way back to your original argument. He's a sly devil, that's for sure...but he adds a lot of quality to our debates here and weeds out the irrational arguments pretty quickly.

That, and he's kinda balding so we can't hurt his feelings too badly...*grin*

James.

A point must be made before it can be refuted and I have seen very little in the way of substantive argument coming from Inquartata. When he does dare to venture into that scary world, as soon as he is challenged he seeks refuge behind a list of logical fallacies. Because of his obsession with Greek rhetoric, I have come to think of him as a Troll in a Toga. That being said, he's obviously an intelligent person and it would be fun to "fence" with him if he would just stay on the friggin strip.

[lochinvar]

Originally Posted by jBirch

How do you figure? This has nothing to do with absolute truth and everything thing to do with probable truth. How is something believed by a very small group of generic people more probably true then something believed by a very large group of generic people?

Do the names "Galileo" and "Copernicus" hold any meaning for you?

[jBirch]

Originally Posted by lochinvar

Do the names "Galileo" and "Copernicus" hold any meaning for you?

A'int they in some fancy book someplace?

James.

[Inquartata]

Originally Posted by jBirch

No, actually, it was your original assertion that needs to be proven. That the more often something is observed to be true has no bearing on whether it is more likely to be true or not.

It's an accepted truism in logic, meaning that it's been proven already. There is no need to reinvent the wheel. Certain basic things are acceptable as assumptions: the Earth orbits the sun, water is wet, Inq fences sabre. These don't have to be proven all over again every time they are cited.

Right. And the more statistical points that support a certain conclusion, the more likely that conclusion is to be the correct one. When P = 1, then you have an absolute truth.

Yes. But this is where you fly off the track: people's opinions or feelings are not "statistical points". They are not facts. Facts cannot be incorrect. Beliefs can be, and often are.

You are confusing the two, I think. Facts inform people's beliefs, but they are not the same thing as facts themselves. Hence gathering data points consisting of what a number of individuals think about X tells us nothing about X, because we don't know whether those data points are reliable or not...

The Law of Large Numbers applies to outcomes involving discrete, verifiable tests, such as coin flips and dice rolls, not to the likelihood of whether facts or opinions are true.

No, but if everyone tells the same lie a thousand times and one guy tells the truth once, the lie is more probably truth then the actual truth is.

Dr. Goebbels, is that you?

Actually, I think you've read "1984" one too many times. Heh.

History is a litany of things that everyone once believed but were later disproved. The nature of truth doesn't change, only it's acceptance. The sun never revolved around the Earth, no matter how many millions of people believed that it did.

Right, but it is less likely to be true if no one affirms it and more likely to be true if everyone does.

One would like to think so, but sadly reality doesn't seem to bear this out. Especially when it comes to nonempirical matters, like we find in most of politics and other praxeological realms.

So then, how does one prove the existence of ghosts?

Empirical, demonstrable physical proof, gathered and assessed by generally accepted scientific methods.

Note that I do not say that ghosts don't exist. We may simply not have the technology needed to record their existence as yet. But the belief of 10,000 people in them is no sort of substitute. At best the question is an open one; belief does not settle it.

I certainly am open to being convinced. So far you've merely engaged in hand waving and counter-factual argument. Give me any example where the increased observation of an event does not go directly to its probability of existing.

Sure.

Many millions of people believe that God exists. Many millions believe he does not. Both groups cannot be right, can they? One or the other must be wrong, despite the numbers involved. One or the other group holds to a conclusion that is not made any more likely by the size of its membership.

When I was studyiing them in university, the professor interpreted "not spitefully or with malice, not to make others seem to be buffoons" to be directing one to use fallacies to construct one's own arguments in a more efficient manner in order to expose the weakness in your opponent's arguments. This was to be done to give your own points more suasion in the debate, not to give your opponent less.

That's a good use for them, but not the only one, and not the only one Aristotle meant them for.

And if one only uses them that way while refusing to point out the fallacies in counterarguments, one will never convince. The fallacious reasoning is simply too convincing to the person not familiar with them, ie most audiences. One places oneself at an insuperable disadvantage...or else one must start advancing appealing fallacious aguments oneself in defense.

Let me quote "Logic: An Introduction", by Robert Churchill, Prof. of Philosophy at George Washington University:

"Expose a fallacious argument if the opportunity arises. When the fallacy is a formal fallacy, the most effective way of exposing its defect is by means of a counterexample. This consists of presenting an argument of exactly the same form that reaches an absurd or plainly false conclusion---a conclusion your opponent will recognize as unjustified.

In the case of an informal fallacy, a more direct approach is needed. Simply say that the argument is unacceptable and, as clearly as possible, explain why."

Also

"It may be advisable to avoid flatly stating that the argument is fallacious because the other person may not understand by 'argument' the linking of premises and conclusion and may think you are trying to prejudice the issue. It may also be better to use the English rather than the Latin names for the fallacies, so as not to appear condescending".

That last is as ar as he goes in advising discretion, and nowhere does his text advise never calling out a fallacy one encounters or restricting their examination to ones own arguments. ( And I take it that we all recognize what an argument is and what 'fallacy' means---a known, accepted trap in logic into which people are wont to fall, rather than an implied insult to the intelligence or underlying rationality of the arguer. Well, I think MOST of us realize that, anyway. I hope so, since even I fall into them occasionally. )

I am saying that the more observations supporting a conclusion, the more likely to be true that conclusion is.

I know, but it still isn't so when the beliefs of human beings constitute the thing being observed. That's what the ad numeram fallacy is all about. You aren't collecting observations of real-world, objective fact, but rather beliefs. Beliefs are formed in many ways, and can be either wrong, right, or partially both. This is no sort of reliable data to use in judging the validity of a conclusion.

Since we are dealing with probability it is important to note that we are talking about probability of truth, not the arbitration of absolute truth

What? Truth is a binary condition. It either is, or it isn't. It does not depend on probablility.

I think I must be misunderstanding what you're saying somehow. :

Since you've posted no source but your own opinion, yeah.

The definitions I posted earlier were not mine. You may have noticed the quote marks around them...

Would you like their source?

It's called an "appeal to authority".

If it's Wikipedia, is it really an "authority"?

Meanwhile, here are Churchill's definitions:

"Fallacy of numbers

Often, the authoritative appeal is to sheer numbers, as if quantity alone were enough to constitute authority. In some versions of the fallacy of numbers, actions or practices of the crowd serve as justification. This version of the fallacy is sometimes called the appeal to common practice. The irrelevance involved in all such fallacies is the claim that simply because some proposition 'p' is....accepted ( as tradition or by people generally ), then there is a reason for accepting 'p'." ---Ellided to omit the discussion of other fallacies he subgroups with this one, eg ad verecundiam.

"The appeal to sympathy is often called an 'appeal to the people', from the Latin ad populam, or an 'appeal to the gallery' because it is widely used by speakers to persuade audiences to respond favorably to them. In making such an appeal, the speaker frequently uses expressions familiar to the audience in an effort to be perceived as an 'insider'---someone who belongs to the group."

In other words, the ad populam is an appeal not to the sheer prevalence of a belief but to its popularity with a certain specific audience being appealed to. ( Admittedly the distinction is a nice one, as in both cases a majority opinion is being drafted into an argument. In the latter case, though, it's on a narrower basis and often involves specialized in-group or emotional wording. )

[Inquartata]

Originally Posted by jBirch

Sure we do! How else are going to train our minds to recognise "dodgy reasoning"? Remember, common sense isn't.

It's also a sort of verbal shorthand. Imagine trying to talk about fencing if instead of using the commonly accepted terminology----lunge, parry, riposte, priority, attack, beat, prise---one had to stop and write a description of each action, every time it occurred...

Much easier just to familiarize oneself with the terminology of logic and then everyone knows what you're talking about.

I'd call it the Law of Parsimony if that wasn't already taken.

[Neil]

My earlier post wasn't directed at Inquartata; I've noticed several people big on quoting fallacies in many forums.

Originally Posted by jBirch

Sure we do! How else are going to train our minds to recognise "dodgy reasoning"? Remember, common sense isn't.

I don't think you need to study the fallacies to be able to recognise poor reasoning. I can recognise various fallacies without knowing what they're called, and my powers of deduction aren't extraordinary. I dont like people using technical terms when they aren't necessary. It needlessly limits the number of people who can follow and participate in the discussion.

Originally Posted by jBirch

It goes to ethos in an argument. The formal purpose of an argument is to persuade everyone that your opinion is correct. You do this by first setting out your own credentials and making a case for why you need to be listened to. This is called Ethos and gives any opinion on the subject at hand almost the weight of fact. Think Expert Testimony in a court of law. This makes it so that people will listen attentively to what you have to say. You then engage the emotions of your audience and get them to suspend their critical judgement so that your ability to launch a new opinion has a receptive audience. This is called Pathos and makes your audience *want* to accept your opinion without necessarily evaluating it too closely. Finally, you logically tie together the emotion and your credibility into the rationale for why your opinion is right. This is called logos and makes it easy for the audience to satisfy their Pathos. They see the lines from the fact to the opinion and accept them as valid. The speech between Mark Antony and Brutus demonstrates this principal admirably.

The ethos/pathos/logos descriptions are interesting, and I haven't heard about these before. I'll have to look up a description. Based on your summary, I agree that ethos and pathos are necessary for a good (rhetorical) argument, but they are only there to support the logos. Without the logos, there is no argument. I think it's annoying when people point out superficial fallacies instead of looking at what the person *meant* and engaging with that.

[Dr. Pfleschbach]

Originally Posted by Inquartata

It's also a sort of verbal shorthand. Imagine trying to talk about fencing if instead of using the commonly accepted terminology----lunge, parry, riposte, priority, attack, beat, prise---one had to stop and write a description of each action, every time it occurred...

Much easier just to familiarize oneself with the terminology of logic and then everyone knows what you're talking about.

I'd call it the Law of Parsimony if that wasn't already taken.

It's certainly fair game to attack fallacies and to use shorthand when doing so, but they should be referenced to specific arguments and a counter argument should still be presented. Otherwise the conversation comes to a screeching halt (maybe that is the goal sometimes). I also think it is constructive for each side to look beyond the presentation, if possible, and try to understand the essence of the idea being expressed.

[jBirch]

Originally Posted by Inquartata

Yes. But this is where you fly off the track: people's opinions or feelings are not "statistical points". They are not facts. Facts cannot be incorrect. Beliefs can be, and often are.

You are confusing the two, I think. Facts inform people's beliefs, but they are not the same thing as facts themselves. Hence gathering data points consisting of what a number of individuals think about X tells us nothing about X, because we don't know whether those data points are reliable or not...

The Law of Large Numbers applies to outcomes involving discrete, verifiable tests, such as coin flips and dice rolls, not to the likelihood of whether facts or opinions are true.

Oh, I see where you're going... My point, is only on the increased or decreased likelyhood of truth. Nothing about the actual underlying truth or not.

The original quote made the comment that a belief is not more likely to be true the more people that believe it. I was refuting that and saying that since it is making a statistical argument that the assertion was flat out wrong by all statistical norms.

By counter-example, The only other argument is that the number of people believing something has no correlation to its likelyhood of truth. Which means that one person believing something is equally likely to be true as 5 billion people believing the same thing. Clearly (and restricting the argument to the likelyhood only) the more people that believe the more likely to be true the argument is (even if it is only fractionally more probable).

History is a litany of things that everyone once believed but were later disproved. The nature of truth doesn't change, only it's acceptance. The sun never revolved around the Earth, no matter how many millions of people believed that it did.

Which has no bearing on the LIKELYHOOD of the opinion being true in the first place.

One would like to think so, but sadly reality doesn't seem to bear this out. Especially when it comes to nonempirical matters, like we find in most of politics and other praxeological realms.

And in the same vein, you can't use logic at all to evaluate nonempirical matters since there is no discrete "truth" to be determined.

Empirical, demonstrable physical proof, gathered and assessed by generally accepted scientific methods.

Which, as a method, fails for random events.

Note that I do not say that ghosts don't exist. We may simply not have the technology needed to record their existence as yet. But the belief of 10,000 people in them is no sort of substitute. At best the question is an open one; belief does not settle it.

I never said it did. But let's suppose that the number 10,000 is inflated to 10,000,000,000. Are ghosts more likely to exist given that number? Especially if those 10,000,000,000 claim to have observed the phenomenon?

Sure.

Many millions of people believe that God exists. Many millions believe he does not. Both groups cannot be right, can they? One or the other must be wrong, despite the numbers involved. One or the other group holds to a conclusion that is not made any more likely by the size of its membership.

In absolute numbers, no. But in relative numbers? If there are a billion people who believe God does not exist and one that believes He does, is God more or less likely to exist?

"It may be advisable to avoid flatly stating that the argument is fallacious because the other person may not understand by 'argument' the linking of premises and conclusion and may think you are trying to prejudice the issue. It may also be better to use the English rather than the Latin names for the fallacies, so as not to appear condescending".

And Neil's (and the Dr's) original point was that the pointing out of fallacies without a valid counter-example is poor rhetorical style. And rude to boot.

What? Truth is a binary condition. It either is, or it isn't. It does not depend on probablility.

I think I must be misunderstanding what you're saying somehow. :

I think you are too. The assertion you presented originally is a statistical-style argument. That means that any conclusion is confined to a probabilty of truth, and not an absolute truth. My point was that the fallacy ad numerum and ad populum only applies to absolute truth arguments and not to statistical type ones.

In this case, TRUTH IS NOT A BINARY.

The definitions I posted earlier were not mine. You may have noticed the quote marks around them...

Would you like their source?

Sure. After-all, you questioned my appeal to authority with a counter-claim to a different authority. *grin*

BTW, *THIS* is what a logical fallacies thread should be about. *grin*

James.

[Inquartata]

Originally Posted by Dr. Pfleschbach

It's certainly fair game to attack fallacies and to use shorthand when doing so, but they should be referenced to specific arguments and a counter argument should still be presented. Otherwise the conversation comes to a screeching halt (maybe that is the goal sometimes).

In many cases I think that would be redundant.

For example, let us say that someone advances the argument that Dick Cheney is in the pocket of Big Oil because of his past employment with Halliburton.

What is the "counterargument" to that? Long and complex, and superfluous, because if the object is to refute the original argument it is enough to show that it doesn't really prove anything---that it's a false argument. Pointing out how and why it is so vitiates its force; that is enough to achieve the refutation.

Of course, in real life that seldom ends the matter. But that's part of what makes debates so much fun.

I also think it is constructive for each side to look beyond the presentation, if possible, and try to understand the essence of the idea being expressed.

Ah, but now we're edging away from Aristotle toward Socrates.

[Inquartata]

Originally Posted by jBirch

The original quote made the comment that a belief is not more likely to be true the more people that believe it. I was refuting that and saying that since it is making a statistical argument that the assertion was flat out wrong by all statistical norms.

Maybe I'm missing some subtle distinction you're seeing, but to me that sounds like the same thing written a different way.

My position, and the position embodied in the idea of the ad numeram fallacy, is that a thing is not made more likely to be true by a large number of people believing it to be true. The objective truth and its acceptance by masses of humanity are not correlated.

And in the same vein, you can't use logic at all to evaluate nonempirical matters since there is no discrete "truth" to be determined.

There IS a truth, in many cases, but there is simply no way to obtain its acceptance by logic because of the idiosyncracies of human perception, thinking and belief.

Which, as a method, fails for random events.

Why? And in what sense is a phenomenon that exists "random"? ( Even discounting chaos theory. )

I never said it did. But let's suppose that the number 10,000 is inflated to 10,000,000,000. Are ghosts more likely to exist given that number? Especially if those 10,000,000,000 claim to have observed the phenomenon?

Nope. That's called "mass hysteria".

How many credulous listeners to the Welles radio broadcast of "War of the Worlds" would have been needed to make an actual Martian invasion "more likely to be true"? How many bullish investors would have been necessary to make the Great Crash never happen...because their mass belief would have made it impossible for it to happen? How many fervent followers of a pyramid scheme would suffice to make the nature of the pyramid into something else, something infinitely sustainable?

In absolute numbers, no. But in relative numbers? If there are a billion people who believe God does not exist and one that believes He does, is God more or less likely to exist?

Neither. He may either exist or not, independently of how many people believe in him. There is no logical connection between the two conditions.

If God is demonstrably in evidence, that will cause people to believe he exists; however, it does not work the other way: no number of people believing God exists will cause him to do so if he does not.

Substitute "any given phenomenon being argued true" for "God" in that statement, and there is why ad numeram is a fallacy.

And Neil's (and the Dr's) original point was that the pointing out of fallacies without a valid counter-example is poor rhetorical style. And rude to boot.

And sometimes necessary. Far more often, the most succinct method.

Counter-examples are not always needed, as an earlier quote specified. Nor are they often effective, because one's opponent can simply contend that the counterexample is a "bad analogy". Then you get bogged down arguing about how it is or isn't...

But an obviously fallacious argument cannot hide itself so easily. As Neil said, its deficiency is apparent to the perspicacious already. But it is not so to all. And least of all to the person who advanced it, or he would not have done so in the first place. It is imperative that it be pointed out to him. That alone will often suffice to refute his argument. To go farther is mere gilding of the lily.

The assertion you presented originally is a statistical-style argument.

A nice trick, since I don't have any idea what a "statistical-style argument" is. Yay, me! I have unplumbed powers!

That means that any conclusion is confined to a probabilty of truth, and not an absolute truth. My point was that the fallacy ad numerum and ad populum only applies to absolute truth arguments and not to statistical type ones.

In this case, TRUTH IS NOT A BINARY.

Truth either is, or is not. There is no more "a little bit true" than there is "a little bit pregnant".

Sure. After-all, you questioned my appeal to authority with a counter-claim to a different authority.

Yes, but YOUR appeal was in itself a counter-claim to:



http://www.infidels.org/news/atheism/logic.html

That's the source of my original definitions. It's the resource to which I usually resort online, when my books are at home and I am not.

[jBirch]

Originally Posted by Inquartata

Maybe I'm missing some subtle distinction you're seeing, but to me that sounds like the same thing written a different way.

My position, and the position embodied in the idea of the ad numeram fallacy, is that a thing is not made more likely to be true by a large number of people believing it to be true. The objective truth and its acceptance by masses of humanity are not correlated.

I am saying the same thing in different ways, absolutely.

The core idea is that a truth doesn't have to be binary especially before it can be measured to be true or not. Start there as that is the essense of a statistical-style argument. My point originally was that ad numerum is a logical fallacy when applied to binary truth tables but not when applied to statistical arguments. The phrase "more likely" marks a statistical argument while the phrase "is true/is not true" marks a binary argument.

So the ad numerum/populum fallacy is that the more people that believe a thing has no bearing on whether it is true or not. However...the more people that believe a thing means that thing is more likely to be true. The difference between those two comments is the core of my point and the nuance in terminology that I was contending with your original definition of ad numerum and ad populum.

Why? And in what sense is a phenomenon that exists "random"? ( Even discounting chaos theory. )

Any event that is measurable but not duplicable.

James.

[Dr. Pfleschbach]

Originally Posted by Inquartata

In many cases I think that would be redundant.

For example, let us say that someone advances the argument that Dick Cheney is in the pocket of Big Oil because of his past employment with Halliburton.

What is the "counterargument" to that? Long and complex, and superfluous, because if the object is to refute the original argument it is enough to show that it doesn't really prove anything---that it's a false argument. Pointing out how and why it is so vitiates its force; that is enough to achieve the refutation.

Merely refuting an argument does not does necessarily disprove the assertion being made. While it may assist in diminishing the credibility of the person making the assertion, the tactic may neglect to point the way to the truth of the matter, in which case a counter argument is called for in order to address the subject directly. One thing to consider is whether the goal of a given debate is to discredit the adversary using one's command of wit and rhetoric, or rather to offer arguments that lead to a greater understanding of the subject being discussed.

[jBirch]

Originally Posted by Dr. Pfleschbach

Merely refuting an argument does not does necessarily disprove the assertion being made. While it may assist in diminishing the credibility of the person making the assertion, the tactic may neglect to point the way to the truth of the matter, in which case a counter argument is called for in order to address the subject directly. One thing to consider is whether the goal of a given debate is to discredit the adversary using one's command of wit and rhetoric, or rather to offer arguments that lead to a greater understanding of the subject being discussed.

Ah, yes, the determination as to whether a discussion is an effort at persuasion or questioning for truth. First thing you have to determine at the outset.

On F.NET sometimes it's a blending of the two. As an individual though, you can take what you want from the discussion without having to commit to one position or another.

James.