I think the question is too broad, isn't it? I mean, you have rules of logic and valid inferences and the like; but it is impossible to generalise as to what will constitute a winning argument in any and all circumstances.

Given that consideration, I would say that a good argument is consistent (is not self-contradictory), logical (does not violate rules of logic), complete (doesn't omit important facts or considerations), and intelligible.

Even so, there will be many questions that may be undecidable, or issues where one might need to take a stand on the basis of intuition or conviction rather than on a clearly articulated rationale; particularly in regards to 'matters of ultimate concern'.

It is easier in matters which are circumscribed; for example, in a chess game, the 'winning argument' is that which wins the game, given the constraints of the game and the rules that govern it; but often, in life, we are confronted with imponderables or unknowables and have to make judgements on the basis of partial knowledge. That is where it is difficult, and why I suspect it will never be possible to come up with a generalised method for 'framing a winning argument' that will apply to all situations.

So you are saying, a good argument may indeed be quite obtuse, as yours appears to be.

I would have thought that as far as criteria go for judging the merit of an argument, 'winning' and 'good' would be at least to some extent synonomous.

alright then, abstruse.

OK I see what you mean. Then in that case, you are saying there must be some criteria other than persuasiveness or efficacy to judge what is a good argument. So presumably this must be something in the area of value judgement? So the question is as much 'what is good' as 'what is a good argument'?

I think the question is too broad, isn't it? I mean, you have rules of logic and valid inferences and the like; but it is impossible to generalise as to what will constitute a winning argument in any and all circumstances.

Given that consideration, I would say that a good argument is consistent (is not self-contradictory), logical (does not violate rules of logic), complete (doesn't omit important facts or considerations), and intelligible.

Even so, there will be many questions that may be undecidable, or issues where one might need to take a stand on the basis of intuition or conviction rather than on a clearly articulated rationale; particularly in regards to 'matters of ultimate concern'.

It is easier in matters which are circumscribed; for example, in a chess game, the 'winning argument' is that which wins the game, given the constraints of the game and the rules that govern it; but often, in life, we are confronted with imponderables or unknowables and have to make judgements on the basis of partial knowledge. That is where it is difficult, and why I suspect it will never be possible to come up with a generalised method for 'framing a winning argument' that will apply to all situations.

But you are assuming that a good argument must be a winning argument, which means what? An argument which convinces its audience? But an argument may be good, but not convince its audience. I don't think that any argument would convince Adminijihad (or Iran) that there was a Holocaust, but does that mean that there is not overwhelming evidence that the Holocaust took place? So what does a winning argument have to do with a good argument? Nothing much, I think.

It seems to me that a good argument is an argument that establishes its conclusion. The term, "good" as my dictionary tells me, is the most general adjective of commendation, so the question is what are the criteria which are necessary and sufficient to commend an argument for establishing its conclusion. As Aristotle pointed out, a good knife would be a knife that cuts well since the function of a knife is to cut. So, if the function of an argument is to establish its conclusion, then a good argument is one that establishes its conclusion well. (By the way, what was abstruse about my argument?)

I think you need to read up on logic before trying to answer this question. Your characterization of a good argument is nonsensical to a person who has studied logic.

Arguments cannot meaningfully be said to be consistent or inconsistent. That doesn't make sense. Do know what inconsistent means? The only interpretation I can come with of your words is that an argument with inconsistent premises is not a good argument. But all such arguments are deductively valid (d. valid) in standard logic. (Because everything follows from a contradiction. This is called The Principle of Explosion.) And what about the form of arguments called Reductio ad Absurdum? They all feature inconsistent premises yet the form is widely (by philosophers) considered a good form of argument.

What does it even mean for an argument to violate a rule of logic? Perhaps you mean that it doesn't have a d. valid form. Though many good arguments do not have a d. valid form either. Not all d. valid arguments have a d. valid form.

If you're interesting in a textbook on logic, just send me a PM. I've got one in PDF.

---------- Post added 11-01-2009 at 04:29 PM ----------



I suppose one could try some ideal observer approach. A good argument is one that will convince an ideally rational reader. (Whatever that means.)

---------- Post added 11-01-2009 at 04:31 PM ----------



But what does it mean to establish a conclusion, if not to convince someone (ideal observer?) of that conclusion? What are the criteria for an argument establishing the conclusion?

It seems to me that a good argument is an argument that establishes its conclusion. The term, "good" as my dictionary tells me, is the most general adjective of commendation, so the question is what are the criteria which are necessary and sufficient to commend an argument for establishing its conclusion. As Aristotle pointed out, a good knife would be a knife that cuts well since the function of a knife is to cut. So, if the function of an argument is to establish its conclusion, then a good argument is one that establishes its conclusion well.

It seems to me that the criterion for establishing a conclusion in the case of deductive arguments is soundness (valid with true premises). A cogent deductive argument would be an argument whose premises are known to be true, and is valid. It is, of course, more complex in the case of non-deductive arguments.

Do we really want to say that all arguments have the same criteria for determining assent? And isn't a part of the assent determined by whether a particular argument is appropriate to the conclusion, and this means: applying another set of standards? What makes a "good" aesthetic argument, a "good" ethical or political argument?

And isn't another part, or aspect, of it somewhat negative in that, assuming the reader is knowledgeable, he decides whether it violates the rules of "formal" logic or commits informal fallacies?

Is the "validity" of a deductive argument the same as the "validity" of an inductive argument? And what about an argument that has a multitude of "steps" that employs both to reach a larger conclusion?

Many arguments are not "arguments" at all, but a detailed description (consider Being and Time one large argument) of a state of affairs, and while we consider each step therein, over and above that we also consider whether, taken as a whole, the description warrants our assent (in the form of a change of perspective, for example). What happens when, again, we accept one or two of the steps, but reject the final conclusion of a long chain of arguments?

So, you are essentially just taking the standard method for evaluating logical arguments (i.e. in the case of deductive logic, validity and soundness), and calling it "good". Ok, this is fine, but I don't get your point. As we know from logic 101, in order to establish a conclusion, we need soundness in the case of a deductive argument. And your argument here just basically amounts to a normative statement saying, "this is good"...

However, you can have sound arguments, that do establish their conclusion well, but beg the question. You also have to consider the context of an argument. If I make a sound argument that establishes its conclusion very well, it might still be "bad" if I'm making this argument about which car-maker is in the best financial shape, when the topic of discussion is on the quality of macrobrews...:bigsmile:



It doesn't just 'seem to you' how we establish a conclusion with deductive arguments; it's standard practice to evaluate for validity and soundness. And 'cogency' is used for evaluating an inductive argument, not a deductive argument.

An argument that is sound, but not about the subject matter, is not a logically bad argument. It is simply irrelevant.

So now we are suddenly dealing with what is 'logically bad' or 'logically good'? You were originally dealing with just 'good' or 'bad'. Of course, a 'logically bad' argument would just be one that is invalid. I could also make a 'logically good' argument that is valid, but unsound. We'd still call that a 'bad' argument though.

The reply to Jeepers is a little harsh, Emil. Some people are not steeped in logic. Although, argument is central to philosophy, so anyone who philosophizes really ought to understand the basics of logic. Otherwise, it is like trying to row a boat without oars.

Yes. We can say that a good argument is one that ought to convince a rational person. That is why Peirce called logic "a normative science".

It seems to me that the criterion for establishing a conclusion in the case of deductive arguments is soundness (valid with true premises). A cogent deductive argument would be an argument whose premises are known to be true, and is valid. It is, of course, more complex in the case of non-deductive arguments.

Do we really want to say that all arguments have the same criteria for determining assent?

Is the "validity" of a deductive argument the same as the "validity" of an inductive argument? And what about an argument that has a multitude of "steps" that employs both to reach a larger conclusion?

Many arguments are not "arguments" at all, but a detailed description (consider Being and Time one large argument) of a state of affairs, and while we consider each step therein, over and above that we also consider whether, taken as a whole, the description warrants our assent (in the form of a change of perspective, for example). What happens when, again, we accept one or two of the steps, but reject the final conclusion of a long chain of arguments?

I attempted to be less harsh that I would normally be. If I found a person arguing about automobile mechanics, I'd expect him to know something about automobile mechanics. Similarly with logic; I expect him to know at least some logic. At the bare minimum a fair grasp of propositional logic is.



We could do many things. Is this how you think one ought to determine whether an argument is good or not, that is, by determining whether it would convince a 'rational person' (vague)?



But what about circular arguments and what about arguments that beg the question? They are often sound, but still they are not good arguments.

"Cogency" is applicable to inductive arguments, not deductive.

If you mean does an argument commit a fallacy or not, then it is not up to anyone to decide that it does, or does not. It is quite objective that, for instance, the fallacy of affirming the consequent, or denying the antecedent, are fallacies. The conclusions of such arguments quite simply fail to follow from their premises. There is nothing subjective about that.

Inductive arguments are neither valid nor invalid. Only deductive arguments are valid or invalid.

Determining whether the premises of a deductive argument establish its conclusion is different from determining whether the premises of an inductive argument establish its conclusion if that is what you mean. Of course, an argument can consist of various sub-arguments (called "lemmas") to reach the final conclusion, if that is what you have in mind. But each argument (lemma) (I am not talking about deductive arguments) must be sound for the entire argument to be sound.

Yes. I simply was saying that in the case of deduction, soundness the criterion for establishing the conclusion of a deductive argument, and thus, a good deductive argument. I was simply trying to say something true, not novel. Soundness is the criterion for a good deductive argument just as keeping the time correctly is the criterion for a good watch. As I said, Peirce pointed out that logic is a normative science, and saying an argument is a good argument is a normative (although not moral) statement just as saying that a watch is a good watch is also a normative (but not moral) statement. Should I not say it even if it is true?

You are right about the fallacy of begging the question. Begging the question is an epistemological, not a logical defect, and it needs special treatment.

Since this was all in the logic forum, I thought that "logic" was understood.

A logically bad (deductive) argument is an unsound argument. A sound argument establishes its conclusion. Since all sound arguments have true conclusions. (Although, of course, it is not true that only sound arguments have true conclusions).

Is this how you think one ought to determine whether an argument is good or not, that is, by determining whether it would convince a 'rational person'

Absolutely not. All I said (in response to your suggestion about idea observers) is that a good argument ought to convince a rational person. Whether or not it does is a different question. We judge a good argument by logical criteria. Just as a necessary condition of a good watch is that it perform its function of keeping time well, so a necessary condition of a good argument is that it perform the function of arguments (to establish their conclusion) well.

Soundness is a necessary condition of good arguments. But as begging the question and circularity show, it is not a sufficient condition of good arguments.

What I mean by a cogent deductive argument is that the premises are known to be true, and the argument is valid. So cogency implies soundness, but not conversely.

Your examples are formal fallacies. I agree that the matter of formal fallacies is objective. but I don't agree that it is objective about all informal fallacies, especially begging the question. You do recall our long discussion on that, right? It's right here.



What makes you think that? I suppose that most inductive arguments are d. invalid. I'm unsure whether all inductive arguments are d. invalid. (Cf. the two views about how to distinguish between inductive and deductive arguments above.)



You say that you're "not talking about deductive arguments", but still say that they "must be sound for the entire argument to be sound.", how does that make sense? Especially since you just said that inductive arguments are not valid or invalid. As you know, soundness implies validity. Your position is incoherent!



You say that "soundness is the criterion for a good deductive argument". Notice the definite article ("the"). Do you mean to say that it is a necessary and sufficient condition? If so, then what about informal fallacies such as circular logic and begging the question?



Do you think it is an objective fact that an argument commits a (formal or informal) fallacy or not? If yes, then you need to believe that whether an argument begs the question is also an objective fact, otherwise your position is inconsistent. Do you?

---------- Post added 11-01-2009 at 09:25 PM ----------



You say that "A logically bad (deductive) argument is an unsound argument". Do you mean that being unsound is a necessary and not-sufficient, not-necessary and sufficient, or necessary and sufficient condition for a deductive argument to be bad?

Your examples are formal fallacies. I agree that the matter of formal fallacies is objective. but I don't agree that it is objective about all informal fallacies, especially begging the question. You do recall our long discussion on that, right? It's right here.

Informal fallacies are fallacies. Fallacious (deductive) arguments (with the exception of begging the question which needs different treatment) are arguments whose premises can be true, but whose conclusion might be false. I still do not think that whether begging the question is fallacious is "subjective" in the sense that if I think an argument commits the fallacy, then it does. But, of course, you may have another meaning of "subjective" in mind. Everybody seems to.

I pointed out that soundness cannot be a sufficient condition of good argument, because of begging the question. It is, however, a necessary condition.

Yes, I think that whether an argument begs the question is an "objective fact" (whatever that means). At least, an argument does not beg the question just because someone thinks it does. If I were dictator of the world (which I should be) I would ban the terms, "subjective" and "objective" on pain of death.