wandering - 04 January 2009 09:30 AM

When someone argues

“X is true because A says so”,

it is usually adressed as “fallacy from authority”.

But we canv view it as an argument with an implicit premise

“A says X
implicit premise: everything A says is true.
X is true”.

And thus the argument is perfectly valid.
....

The crux of my problem is that an argument that fails for a reason of logic, and one that fails for reason of soundness are two very different types of an argument. However, by merely reformulating it, I changed its nature completely.

Hmmm. The “argument from authority” is already an informal fallacy not a formal fallacy of logic - that is these are all questions of sound premises not logical validity, so I do not see what you have changed here at all.

dougsmith - 04 January 2009 09:50 AM

wandering - 04 January 2009 09:30 AM

The crux of my problem is that an argument that fails for a reason of logic, and one that fails for reason of soundness are two very different types of an argument. However, by merely reformulating it, I changed its nature completely.

OK.

No, it’s not ok. One shouldn’t be able to change the nature of the fallacy by reformulating the argument.

Bryan - 04 January 2009 09:46 AM

Whoa, there.  You wrote hastily, for there is no barrier to creating an invalid argument for the spheroid shape of the Earth, including one that contains the true premise that the earth is spheroid in shape.  I have little doubt that you agree with that, but your sentence conveys the opposite impression.

Yeah, I realized that as I was writing it. I’m writing quickly and eliding some steps in the argument. What I should have said is that one could have written a valid and sound argument (e.g. that you can get to India by going west across the Atlantic Ocean) with one premise that the earth is round. But that argument would have “begged the question” at the time.

So a valid and sound argument can “beg the question”. That’s the point I was making.

Bryan - 04 January 2009 09:46 AM

In my opinion, the error in reasoning is implicit in the presentation of an argument that is intended to convince others of the conclusion.  It is not reasonable to convince somebody of the truth of a conclusion by using that very conclusion as a premise in a circular argument recommending the conclusion in turn.

Well, but as you know, any valid argument has the conclusion implicit in the premises. It may not be obviously implicit, but logically the conclusion cannot contain information that wasn’t in the premises already. This is the problem with validity and “begging the question”.

That said, I do agree with you that there is something wrong with many arguments that “beg the question”. The problem is not that they are invalid (let’s assume the arguments presented are always valid) but that the premises are unsupported. Clearly, an argument that is intended to convince others must do so in part by convincing them that the premises are true. But not everyone will be convinceable in this way—indeed, I would be willing to argue that there is no possible argument which will be always convincing. Similarly, there is no premise that everyone will take as true. So given any set of premises, some people will claim that those premises in that argument “beg the question” in themselves being unsupported.

Again, I’m not claiming that there’s nothing wrong with begging the question. What I would claim is that sometimes begging the question is a problem (where the premises are amenable to convincing counterargument) and sometimes begging the question is not a problem (where the premises are amenable to unconvincing counterargument).

Of course, what counts as convincing and unconvincing will depend on the person doing the counting.

wandering - 04 January 2009 09:53 AM

No, it’s not ok. One shouldn’t be able to change the nature of the fallacy by reformulating the argument.

Why not?

Fallacies are implicitly dependent on the form of the argument. Reformulate the argument, you (could) get a different fallacy.

faithlessgod - 04 January 2009 09:53 AM

Hmmm. The “argument from authority” is already an informal fallacy not a formal fallacy of logic - that is these are all questions of sound premises not logical validity, so I do not see what you have changed here at all.

Right, this is true too, in a sense. But formally this changes a little from an “argument from authority” to an argument with one false premise.

One could also say that every argument from authority is implicitly an argument with one false premise ... but the issue then would be what that premise actually says. (It doesn’t need to be quite as strong as put there).

dougsmith - 04 January 2009 10:02 AM

So a valid and sound argument can “beg the question”. That’s the point I was making.

And that’s a good point so far as it goes, but one of the main measures of an argument is whether it works to get others to accept the conclusion.  Valid and sound arguments that work are preferable; some such offer the audience no reason to accept the conclusion.  I like sound arguments featuring a logical inference apart from merely the principle of identity.

Bryan - 04 January 2009 09:46 AM

In my opinion, the error in reasoning is implicit in the presentation of an argument that is intended to convince others of the conclusion.  It is not reasonable to convince somebody of the truth of a conclusion by using that very conclusion as a premise in a circular argument recommending the conclusion in turn.

Well, but as you know, any valid argument has the conclusion implicit in the premises.

Right, but typically not in one premise, and it is that condition that results in the fallacy of begging the question.

It may not be obviously implicit, but logically the conclusion cannot contain information that wasn’t in the premises already. This is the problem with validity and “begging the question”.

That problem is resolved when one distinguishes between an argument that takes truths and through logical inference produces a conclusion that is different from either/any of the premises alone.

That said, I do agree with you that there is something wrong with many arguments that “beg the question”. The problem is not that they are invalid (let’s assume the arguments presented are always valid) but that the premises are unsupported. Clearly, an argument that is intended to convince others must do so in part by convincing them that the premises are true. But not everyone will be convinceable in this way—indeed, I would be willing to argue that there is no possible argument which will be always convincing.

Granted; there is such a thing as “invincible ignorance.”

Similarly, there is no premise that everyone will take as true. So given any set of premises, some people will claim that those premises in that argument “beg the question” in themselves being unsupported.

I must point out that the fallacy of begging the question applies in particular to the point at issue.  For example, if we argue whether incompatibilist free will can exist coherently in a given scenario (for the sake of argument), then we fallaciously beg the question by assuming causal determinism in a premise.  A reductio ad absurdum that misrepresents the issue doesn’t help us, after all.  We might “beg the question” by assuming that LFW exists as we attempt to reduce it to absurdity, but the attempt requires us to make the assumption for the sake of argument.  An unsupported premise is not the same as the fallacy of begging the question (see “fallacy of the contested premise” for the distinction).

Again, I’m not claiming that there’s nothing wrong with begging the question. What I would claim is that sometimes begging the question is a problem (where the premises are amenable to convincing counterargument) and sometimes begging the question is not a problem (where the premises are amenable to unconvincing counterargument).

I think you need to focus more on the way arguments focus on particular points of contention.  Suppose two folks think the Earth is flat but want to argue whether it would be at least consistent to imagine a spheroid Earth.  Neither accepts the premise that the Earth is flat; neither would regard their arguments as sound.  In that sense, they both “beg the question”—but the relevant sense of the fallacy of begging the question occurs if the argument somehow assumes its conclusion in one of the premises.  That is, one is asking the other to concede the point at issue from the outset rather than convincing him through customary logical inference.

That is the problem with begging the question.

Of course, what counts as convincing and unconvincing will depend on the person doing the counting.

Indeed.  What would advertisers and lawyers do if fallacious arguments lost their ability to convince others?

dougsmith - 04 January 2009 10:04 AM

wandering - 04 January 2009 09:53 AM

No, it’s not ok. One shouldn’t be able to change the nature of the fallacy by reformulating the argument.

Why not?

Fallacies are implicitly dependent on the form of the argument. Reformulate the argument, you (could) get a different fallacy.

The point of a logical fallacy is that it enables you to dismiss the argument at hand. No matter if the premises are true or not, the argument is invalid. You don’t need to think about the external world to do it.

With a statement, even as strong is “A is right all the time”, it is a question of empirical fact, which needs investigation.

Two different issues.

(Though perhaps I could say that a statement as strong as “A is right all the time” is wrong for logical - it is possible to mentally construct a world in which A says something wrong, therefore this statement can’t be right).

dougsmith - 04 January 2009 09:18 AM

Well, but in putting it that way you are introducing the additional concept of something’s “being established”, where this is clearly more than merely a matter of validity or soundness, since it is a property of premises in isolation.

But being established simply means “The conclusion of another sound argument”, so there is no new concept after all.

So an informal a fallacy according to you is such a one that renders the argument invalid because of premises in isolation?

Premises in isolation are possible conclusions of other arguments, so I do not see the importance of your distinction.

wandering - 05 January 2009 03:21 AM

The point of a logical fallacy is that it enables you to dismiss the argument at hand. No matter if the premises are true or not, the argument is invalid. You don’t need to think about the external world to do it.

Yes but your argument from authority was never a logical fallacy it was an informal fallacy - which usually, although not always (Doug?), can be reformulated as unsound premise(s). (And Doug is right too, reformulating or adding in hidden premises can change or add to the fallacies being committed).

wandering - 05 January 2009 03:21 AM

With a statement, even as strong is “A is right all the time”, it is a question of empirical fact, which needs investigation.

This is about sound premises so nothing has changed. The premise above by default is unsound unless you can provide convincing argument otherwise, the burden of proof is on you to show this?

wandering - 05 January 2009 03:35 AM

But being established simply means “The conclusion of another sound argument”, so there is no new concept after all.

Not necessarily. As we’d discussed before, this is only true if you view argumentation as potentially infinite in depth. If there are self-evident propositions, they do not need additional arguments to be established.

Further, when we ask whether an argument is valid and/or fallacious, we are looking at a finite set of premises and a conclusion. We aren’t taking into account the arguments (potential or not) for the premises of that argument.

wandering - 05 January 2009 03:35 AM

So an informal a fallacy according to you is such a one that renders the argument invalid because of premises in isolation?

No, the argument isn’t made invalid by committing an informal fallacy. If it did, the fallacy would be a formal fallacy. If you like, it renders the argument potentially unsound. That is, since the premises are not themselves established as sound, neither is the conclusion, even though the argument is valid.

dougsmith - 05 January 2009 05:18 AM

No, the argument isn’t made invalid by committing an informal fallacy. If it did, the fallacy would be a formal fallacy. If you like, it renders the argument potentially unsound. That is, since the premises are not themselves established as sound, neither is the conclusion, even though the argument is valid.

Yes and potentially unsound that is better than my by default unsound

dougsmith - 05 January 2009 05:18 AM

No, the argument isn’t made invalid by committing an informal fallacy. If it did, the fallacy would be a formal fallacy. If you like, it renders the argument potentially unsound. That is, since the premises are not themselves established as sound, neither is the conclusion, even though the argument is valid.

I think most informal fallacies are also formal fallacies.

For example

“My preacher said X
...
X is true”

Isn’t a valid argument structure.

Or

“You are a horrible person
You believe X
...
X is not true”

As well. If you symbolize this, there is just no relationship between the premises and the conclusion. Zero. So I think they are formal fallacies as well.

=========================

Besides, informal fallacies are about logic, not about potential soundness.
“My Rabbi said X
...
X is true”
can have a true premise, a true conclusion, and yet be an informal fallacy. This proves the problem is with the logic, not with the soundness.

That depends how you opt to formalize them.

dougsmith - 05 January 2009 07:34 AM

That depends how you opt to formalize them.

What do you think is the better way ?

—-

I think adding the additional premise is the better one, since it emphasizes the psychology of the reason error. For example if you formalize an ad populum as

“The majority says X
The majority is always right
—
X is true”

the psychology of the person who says so is very evident.

It can be, but it can also devolve into a straw man argument. E.g., with the fallacy from authority, one can use an authority figure as a premise in an argument without claiming that literally everything that authority figures say is true.