Comment by coldtea | Hacker News Reader

Comment by coldtea | Hacker News Reader

"Your Logical Fallacy Is" giving too much credit to "logical fallacies".

Conversations are not formal proof, in pure logic.

And logical fallacies are not pure logic either. More like guidelines.

In the real world, where you don't have all the information at hand, and you have trust levels and possibilities and resource limits, a "logical fallacy" can be more beneficial (statistically) than not.

Case in point: "appeal to authority". If I'm sick, I'd rather trust my doctor, than a non-doctor guy, just for the fact that the doctor has a MD and he does not. I don't have time to study medicine and examine both approaches (e.g I could be bleeding to death). So the more rational thing to do here is listen to the authority, not the random guy.

Similar issues can be found for most of the other fallacies, from ad hominem to no-true-scotchman and beyond.

Nice guidelines, but by no means complete and by no means appropriate for all real life situations.

Like Occam's Razor. It's just a guideline too. Sometimes the MORE complex story, with the more entities involved in it, is the one that actually happened. E.g in police cases. There is often a clean-cut explanation that might seem obvious but is wrong, and a more involved story that is what actually happened. Lots of innocent have fell prey to the "obvious" and "simple" explanation, only to be proven innocent many years later.

Think of the movie "Arlington Road". What would most people think of the events, not knowing the whole story? "But most real cases are not like Arlington Road!" My point exactly. Most, not all. Occam's razor is a guideline, not a logical necessity.

A large problem with using "cheat sheet" lists of fallacies is that they lack the nuance of the actual fallacies they cite. An argument from authority is fallacious if, and only if, the authority cited is not an unbiased expert in the matter or there is no expert consensus on the matter. Trusting your doctor when you are sick is not an instance of fallacious reasoning.

While I'm here, another problem with things like this is the tendency for people to use them as an argument-refuter-o-matic. Indicating an argument is fallacious is not a rebuttal of the argument: it's merely an indicator that the argument is incomplete. Using fallacies to rebut an argument is, somewhat ironically, itself a fallacy.

"Trusting your doctor when you are sick is not an instance of fallacious reasoning."

Nor is it an argument. It's an action based on a personal risk assessment.

If you said "The cancer studies are trash because my doctor says they are." Then you would be making a fallacious argument. Authority is only rational as a screening tool in the evaluation of large amounts of after information. It most definitely is not a proof.

> Nor is it an argument. It's an action based on a personal risk assessment.

Concluding that something is the best course of action because an expert authority is, in fact, an example of the results of an inductive argument.

    P1: I am sick.
    P2: I have no knowledge of how to treat my sickness.
    P3: Knowing how to treat my sickness requires expert-level knowledge.
    P4: My doctor says I should take X to treat my sickness.
    P5: If my doctor is an unbiased expert whose opinion is representative
        of the medical community on my sickness, then what my doctor is
        prescribing is likely correct.
    P6: I have no reason to believe my doctor is not an unbiased expert
        whose opinion is representative of the medical community on my
        sickness.
    ------------------------------
    C:  If I follow my doctor's prescription, my sickness will most
        likely be treated correctly.

This is a valid argument, and not an appeal to or argument from authority. The conclusion is justified based on the premises provided.

One could, of course, discount any of the premises provided: for example, I could later find out that my doctor is a hack. If I do not then revise my conclusion or introduce new evidence to, then the conclusion would no longer be justified and the argument would become invalid.

> If you said "The cancer studies are trash because my doctor says they are." Then you would be making a fallacious argument.

If you make the claim that certain cancer studies are trash because your doctor says they are despite expert consensus indicating that they are, in fact, conclusive, then yes, your argument would be fallacious.

If, on the other hand, you make the claim that certain cancer studies are trash because your doctor says they are and it so happens that your doctor's position is representative of the oncology community as a whole, your argument would not be fallacious (or, rather, it would not be fallacious merely on the basis of an argument from authority).

> It most definitely is not a proof.

Informal fallacies, like the argument from authority, are indications of faults in an argument: they are not proof-enders, nor do they apply to proofs. So I'm not sure what point you're trying to make by saying an inductive argument is not a proof. No one could, or should, dispute that.

>An argument from authority is fallacious if, and only if, the authority cited is not an unbiased expert in the matter or there is no expert consensus on the matter.

I'm not sure the "appeal to authority" is about that though. If it was meant as you describe it, it could just be called "appeal to a biased authority" or "appeal to a non-authority" and everyone would agree it is fallacious to do so.

Take your example: "Trusting your doctor when you are sick is not an instance of fallacious reasoning".

According to the "fallacy" people, this IS actually a case of fallacious reasoning. It amounts to saying:

"His opinion on this medical matter is correct because he is a doctor"

whereas (they maintain), his opinion could just as well be wrong, and you can't tell without examining the individual specific case thoroughly, through actual examination, not quoting past expertise.

According to the "fallacy" people? What? There is no fallacy standards body that arbitrates normative definitions of fallacies. Fallacies are what they are because they describe faults in reasoning: they describe the terms under which a conclusion does not follow from the premises given.

The argument from authority, argumentum ad auctoritatem, is a well-known and well-understood fallacy. Take the GGP's scenario: I am sick, and there is a doctor who has prescribed treatment or a cure.

By virtue of being a doctor, it's implied that they are a medical expert. Going to a doctor and trusting him is not the same as going to Joe the Plumber and asking him for medical advice.

So if I, being a layman, believe that the doctor is actually a doctor (and thus, an expert) and I have no contrary knowledge to refute the doctor's claim, there is nothing that prevents me from trusting his prescription provided I have no reason to believe he is biased or his prescription goes against the consensus of the greater medical community on my sickness.

That last part is key, and how trusting a doctor when you are sick is not an appeal to authority. If the doctor in question is an expert in the field and expert consensus agrees on a treatment or cure, I can get a second, or third, or fourth opinion from other expert doctors and get the same answer.

So by trusting a doctor, I'm not willy-nilly putting faith in someone else. Rather, I'm using inductive reasoning to make a reasonably valid conclusion.

One could refute any of the premises—that the doctor isn't really an expert, or that there really isn't consensus, or that the doctor isn't really impartial—and that's fine, but if I have every reason to believe that the doctor is, in fact, an unbiased expert and representative of consensus (or, conversely, no reason to believe that he isn't any of those things), I am justified in believing what he says.

>According to the "fallacy" people? What? There is no fallacy standards body that arbitrates normative definitions of fallacies.

Only there is.

The fallacies are what they are because people listed them, gave them fancy latin names, put them up on the internet, and argue about them.

Average Joe doesn't use fallacy listings. Neither he maintains them. The "fallacy" thing is mostly confined in the "sceptic" / "more logical than thou" communities on the intertubes. (Of course the starting origin of the fallacies is in actual philosophers and logicians).

As for the "argument from authority", Wikipedia backs me up on this one (If I'm allowed to appeal to its authority):

>Fallacious arguments from authority often are the result of failing to meet at least one of the required two conditions (legitimate expertise and expert consensus) structurally required in the forms of a statistical syllogism.[1][2] First, when the inference fails to meet the first condition (inexpert authority), it is an appeal to inappropriate authority, which occurs when an inference relies upon a person or a group without relevant expertise or knowledge of the subject matter under discussion.

This is what I said exactly, that in that case you do have a fallacy, but it's actually "appeal to non-authority" (in Wikipedia: "to inappropriate authority").

>Second, because the argument from authority is an inductive-reasoning argument — wherein is implied that the truth of the conclusion cannot be guaranteed by the truth of the premises — it also is fallacious to assert that the conclusion must be true.[2] Such a determinative assertion is a logical non sequitur, because, although the inductive argument might have merit — either probabilistic or statistical — the conclusion does not follow unconditionally, in the sense of being logically necessary.

This is the gist of the other thing I said. Even if the criteria of expertise, non bias, consensus, etc, are met, it is still a fallacy to think that the appeal to authority necessary implies truth.

E.g in your example: X is a doctor. X is unbiased. X believes Y on the matter. The medical consensus is Y on the matter. Therefore Y is true.

Is a fallacy. Consider how the medical community thought of a lot of BS in the past as true, which later was found wrong.

It's important to note the difference between formal and informal fallacies. Formal fallacies are indeed pure logic. Informal fallacies are context-dependent. Appeals to authority and ad hominems are examples of informal fallacies; neither is intrinsically fallacious.

Reply on news.ycombinator.com