I might be mistaken here, but my understanding of informal logical fallacies is that they're useful in judging one's own reasoning to be sound (should one actually care about such things), and not intended to derive a conclusion of the particular argument's falsehood. The position held can be true despite faulty reasoning.
Yes, I think they are supposed to be rules of thumb. But part of my point is that they are not even very useful as rules of thumb.
Take the ad hominem. As I illustrated, there are certain cases where you *are* justified in taking the person into account. Now, if you just plainly apply it, you will get a lot of cases wrong. But if you don't, then you will have to consider whether it applies in any case, and then you are just better of dropping the step where you use the fallacy, and just thinking about whether the person is relevant in this case. So it is either a bad way of reasoning, or a useless extra step.
Okay, but with the Ad Hominem fallacy, it's useful to define the notion that attacking the person or their beliefs does nothing to further whatever point you're attempting to make. Same with a Straw Man defining that someone has strayed off to attacking a point of view that the other person doesn't even hold, it's just one they can prop up for easy defeat.
Defining logical fallacies in general provides a consistent framework/guide for keeping debate/opinion/argument in check from bad-faith claims, desperate emotionally-charged rhetoric, bullying and the desire to just shut down dialog.
Yes, perhaps in these sense that fallacies are rough guidelines for how to be respectful and charitable in a debate, they may have a value. But then that is not how they are usually presented. Rather they are presented as rules for reasoning.
As someone who has taught this stuff many times, the points you are making here are well taken (I particularly liked the discussion of how absence of evidence is evidence of absence). But I think you are overlooking some considerations that make learning and using fallacies useful. It has to do with, as you said, pattern recognition, and also with rhetoric.
I explain fallacies to my students as arguments that can be rhetorically persuasive despite being weak (in the technical, inductive sense). An argument can be persuasive disproportionate to what its effect on your credence rationally should be. Fallacies are important to know because there are certain patterns in the way people who don’t actually have evidence on their side or who don’t know what they are talking can be effective rhetorically. I teach my students that almost every kind of fallacy has a structure that can in some cases be strong, but they are often used despite being weak.
Take ad hominem. Sure, there are times when knowing a person’s character can be important to evaluating their argument. But often it is just very minimally relevant, and it is generally much mire useful to pay attention to what they are saying and judge it on the merits. And once you can recognise the extent to which, say, the average political opinion column is composed of a lengthy string of snide dismissals and insults of the opposing leader, substituting that for any real consideration of what they really stand for, you have learned something useful.
I think those are some very good points! I like the approach of seeing them as things that may have a tendency to raise/lower your credence in something to an unreasonable extent, even if it often should have at least some effect. Basically, they should perhaps be presented as common pitfalls in the presentation of an argument rather than faults in the argument itself. The problem is that the way they are usually presented and used (especially on the internet) is as a list of things which, when part of an argument, automatically make it bad (or at least worse), which I think is just false for most of them.
“For example, an argument which begs the question will always be logically valid (the premises cannot be true without the conclusion being true, since the conclusion is one of the premises”
This strikes me as a little strange. I mean typically there is a distinction made between arguments and use of arguments. To see whether someone is begging the question, you have to focus on the latter; that is, someone begs the question if the reason they give in support of a premise presupposes the truth of the conclusion. This is different from requiring that the conclusion be one of the premises. As a result you can get arguments that look as though they beg the question, but needn’t necessarily beg the question.
For an example of this coming into play, there is a paper by Christopher Peacocke (I think it’s called Descartes defended) that constructs a version of the Cogito which is supposed to avoid begging the question.
I’m with you but I’d go even further. There’s really only one logical fallacy: ambiguity/equivocation. Begging the question (circularity) isn’t a fallacy because every deductively valid argument is circular. Then there’s the general fallacy of deductive argument: that such arguments are really impossibility proofs in disguise.
I don't think every deductive argument is circular. At least, we can come up with a plausible definition of circularity on which it is not the case:
An argument is circular IFF one of the premises of the argument singlehandedly entails the conclusion.
On this definition the following argument would be circular:
1. Socrates is mortal
2. Therefore Socrates is mortal
But the following one is not circular:
1. All men are mortal
2. Socrates is a man
3. Therefore Socrates is mortal
It is not circular because any one of the premises on its own could be true and the conclusion false, but it is valid because all the premises cannot be true together and the conclusion false.
This might just be a verbal dispute, but I don't think it matters with respect to circularity whether the conclusion is confined to one premise or spread out over several. What matters is that the conclusion is already contained in the premises; the conclusion is just the premises rearranged or put together. Circularity is just a matter of the conclusion functioning as a restatement of the premises, adding no new information.
That said, I do think that circular arguments can be informative because one might not have realized that the conclusion could be thusly decomposed.
I think you are right that it is just a verbal dispute, but (unsurprisingly) I also think my definition is more useful;)
After all, basically no circular arguments on my definition are informative or persuasive. But many non-circular valid arguments *are*, since they can show entailments between different beliefs, which you didn't realize were there. Your definition seems to miss this distinction.
As I have expressed elsewhere, I don't think there is a *right* answer to these sorts of things, but I think we can decide a definition which we find useful.
Yeah, but on your definition the following argument would be circular:
(1) A is B and C.
(2) Therefore, A is B and C.
whereas the following argument wouldn't be circular:
(1) A is B
(2) A is C
(3) Therefore, A is B and C.
But the only difference is how the premises are presented. And so, on your definition, circularity isn't a function of an argument's form or content but on the manner of presentation. Circularity can be avoided just by breaking up one premise into two, even if no new information is added.
I don't think your example is super counter-intuitive after all the former is literally just the conclusion twice, while the other brings two considerations together for a conclusion.
Of course for any deductive argument, you could just reformulate it as one premise, which is the conjunction of all the premises, and a conclusion. This would mean any argument could be rewritten to be circular. But I don't think that is too problematic. After all the problem with circular arguments has nothing to do with the form but everything to do with persuasiveness. And it doesn't seem to me that making one long conjunction will be a very persuasive way of presenting an argument. Though this dispute is probably more a matter of personal temperament than anything.
But how much more persuasive can you get than a circular argument? If you accept the premises then you're bound to accept the conclusion :)
Also, your point about expressing all the premises as one long conjunction is important. It shows that a deductive argunent's circularity depends not on its content or form or persuasiveness but on how its premises are individuated.
But, craziest of all, we're arguing about the wrong thing. Who cares how you define circularity? Deductive argument seems capable of generating real insight, in some sense transcending its premises. This is the fabled synthetic a priori. And, somehow, deductive reasoning seems capable of giving it to us. But how?
I think it depends on how you construe it. You may look at it in the sense that them being the CEO of a tobacco-company is strong evidence that they have a character such that they are generally unreliable on topics like these. Though that may be a stretch. It would perhaps have been better to just take an example of something like a known liar vs. someone who is generally trustworthy.
I agree that if everyone just believed something because everyone else did, then upon finding that out, they would have a defeater for that belief. But that is not usually how things work. Rather there are at least a few people who actually look at the evidence for something/think about the argument, and generally there is not reason to think that they are mistaken, or that people reporting that stuff are mistaken or anything like that - especially if it is quite mundane. There can of course also be stuff about the subject matter which provides defeaters, like if it is very controversial, or includes incentives for most people, such as in politics. But again that is a special case, and I don't think it threatens the general thrust.
I might be mistaken here, but my understanding of informal logical fallacies is that they're useful in judging one's own reasoning to be sound (should one actually care about such things), and not intended to derive a conclusion of the particular argument's falsehood. The position held can be true despite faulty reasoning.
Yes, I think they are supposed to be rules of thumb. But part of my point is that they are not even very useful as rules of thumb.
Take the ad hominem. As I illustrated, there are certain cases where you *are* justified in taking the person into account. Now, if you just plainly apply it, you will get a lot of cases wrong. But if you don't, then you will have to consider whether it applies in any case, and then you are just better of dropping the step where you use the fallacy, and just thinking about whether the person is relevant in this case. So it is either a bad way of reasoning, or a useless extra step.
Okay, but with the Ad Hominem fallacy, it's useful to define the notion that attacking the person or their beliefs does nothing to further whatever point you're attempting to make. Same with a Straw Man defining that someone has strayed off to attacking a point of view that the other person doesn't even hold, it's just one they can prop up for easy defeat.
Defining logical fallacies in general provides a consistent framework/guide for keeping debate/opinion/argument in check from bad-faith claims, desperate emotionally-charged rhetoric, bullying and the desire to just shut down dialog.
Yes, perhaps in these sense that fallacies are rough guidelines for how to be respectful and charitable in a debate, they may have a value. But then that is not how they are usually presented. Rather they are presented as rules for reasoning.
As someone who has taught this stuff many times, the points you are making here are well taken (I particularly liked the discussion of how absence of evidence is evidence of absence). But I think you are overlooking some considerations that make learning and using fallacies useful. It has to do with, as you said, pattern recognition, and also with rhetoric.
I explain fallacies to my students as arguments that can be rhetorically persuasive despite being weak (in the technical, inductive sense). An argument can be persuasive disproportionate to what its effect on your credence rationally should be. Fallacies are important to know because there are certain patterns in the way people who don’t actually have evidence on their side or who don’t know what they are talking can be effective rhetorically. I teach my students that almost every kind of fallacy has a structure that can in some cases be strong, but they are often used despite being weak.
Take ad hominem. Sure, there are times when knowing a person’s character can be important to evaluating their argument. But often it is just very minimally relevant, and it is generally much mire useful to pay attention to what they are saying and judge it on the merits. And once you can recognise the extent to which, say, the average political opinion column is composed of a lengthy string of snide dismissals and insults of the opposing leader, substituting that for any real consideration of what they really stand for, you have learned something useful.
I think those are some very good points! I like the approach of seeing them as things that may have a tendency to raise/lower your credence in something to an unreasonable extent, even if it often should have at least some effect. Basically, they should perhaps be presented as common pitfalls in the presentation of an argument rather than faults in the argument itself. The problem is that the way they are usually presented and used (especially on the internet) is as a list of things which, when part of an argument, automatically make it bad (or at least worse), which I think is just false for most of them.
Tha hard on everyone has for fallacies on Reddit is tiresome lol
“For example, an argument which begs the question will always be logically valid (the premises cannot be true without the conclusion being true, since the conclusion is one of the premises”
This strikes me as a little strange. I mean typically there is a distinction made between arguments and use of arguments. To see whether someone is begging the question, you have to focus on the latter; that is, someone begs the question if the reason they give in support of a premise presupposes the truth of the conclusion. This is different from requiring that the conclusion be one of the premises. As a result you can get arguments that look as though they beg the question, but needn’t necessarily beg the question.
For an example of this coming into play, there is a paper by Christopher Peacocke (I think it’s called Descartes defended) that constructs a version of the Cogito which is supposed to avoid begging the question.
God! You know First Order Logic even before starting uni, wow! Quite impressive
Hehe, thank you! Though I wouldn't say I *know* it - more like can use it in emergencies:)
I’m with you but I’d go even further. There’s really only one logical fallacy: ambiguity/equivocation. Begging the question (circularity) isn’t a fallacy because every deductively valid argument is circular. Then there’s the general fallacy of deductive argument: that such arguments are really impossibility proofs in disguise.
I don't think every deductive argument is circular. At least, we can come up with a plausible definition of circularity on which it is not the case:
An argument is circular IFF one of the premises of the argument singlehandedly entails the conclusion.
On this definition the following argument would be circular:
1. Socrates is mortal
2. Therefore Socrates is mortal
But the following one is not circular:
1. All men are mortal
2. Socrates is a man
3. Therefore Socrates is mortal
It is not circular because any one of the premises on its own could be true and the conclusion false, but it is valid because all the premises cannot be true together and the conclusion false.
This might just be a verbal dispute, but I don't think it matters with respect to circularity whether the conclusion is confined to one premise or spread out over several. What matters is that the conclusion is already contained in the premises; the conclusion is just the premises rearranged or put together. Circularity is just a matter of the conclusion functioning as a restatement of the premises, adding no new information.
That said, I do think that circular arguments can be informative because one might not have realized that the conclusion could be thusly decomposed.
I think you are right that it is just a verbal dispute, but (unsurprisingly) I also think my definition is more useful;)
After all, basically no circular arguments on my definition are informative or persuasive. But many non-circular valid arguments *are*, since they can show entailments between different beliefs, which you didn't realize were there. Your definition seems to miss this distinction.
As I have expressed elsewhere, I don't think there is a *right* answer to these sorts of things, but I think we can decide a definition which we find useful.
Yeah, but on your definition the following argument would be circular:
(1) A is B and C.
(2) Therefore, A is B and C.
whereas the following argument wouldn't be circular:
(1) A is B
(2) A is C
(3) Therefore, A is B and C.
But the only difference is how the premises are presented. And so, on your definition, circularity isn't a function of an argument's form or content but on the manner of presentation. Circularity can be avoided just by breaking up one premise into two, even if no new information is added.
I don't think your example is super counter-intuitive after all the former is literally just the conclusion twice, while the other brings two considerations together for a conclusion.
Of course for any deductive argument, you could just reformulate it as one premise, which is the conjunction of all the premises, and a conclusion. This would mean any argument could be rewritten to be circular. But I don't think that is too problematic. After all the problem with circular arguments has nothing to do with the form but everything to do with persuasiveness. And it doesn't seem to me that making one long conjunction will be a very persuasive way of presenting an argument. Though this dispute is probably more a matter of personal temperament than anything.
But how much more persuasive can you get than a circular argument? If you accept the premises then you're bound to accept the conclusion :)
Also, your point about expressing all the premises as one long conjunction is important. It shows that a deductive argunent's circularity depends not on its content or form or persuasiveness but on how its premises are individuated.
But, craziest of all, we're arguing about the wrong thing. Who cares how you define circularity? Deductive argument seems capable of generating real insight, in some sense transcending its premises. This is the fabled synthetic a priori. And, somehow, deductive reasoning seems capable of giving it to us. But how?
I think it depends on how you construe it. You may look at it in the sense that them being the CEO of a tobacco-company is strong evidence that they have a character such that they are generally unreliable on topics like these. Though that may be a stretch. It would perhaps have been better to just take an example of something like a known liar vs. someone who is generally trustworthy.
I agree that if everyone just believed something because everyone else did, then upon finding that out, they would have a defeater for that belief. But that is not usually how things work. Rather there are at least a few people who actually look at the evidence for something/think about the argument, and generally there is not reason to think that they are mistaken, or that people reporting that stuff are mistaken or anything like that - especially if it is quite mundane. There can of course also be stuff about the subject matter which provides defeaters, like if it is very controversial, or includes incentives for most people, such as in politics. But again that is a special case, and I don't think it threatens the general thrust.