Comment by pessimizer | Hacker News Reader

Comment by pessimizer | Hacker News Reader

Parent Full thread

>There's almost always a good argument for any given thing, and often several.

And if there aren't better arguments for the other things, then those are the things that should be done. If you're telling me that there are almost always equally good arguments for every possible choice in all decisions that are made, I'm disagreeing with that. If that's not what you're saying, I don't know what you're saying.

>It's sometimes useful to be able to say "yes, there is a huge pile of good arguments in favor of X here. And I don't have time to address each one. But it's important to note that the author of this pile of good arguments stands to make a great deal of money if people believe X, regardless of its truth."

It may be important to note, but it's not argumentation. If it's a good argument, accept that it's a good argument and move on. Just because people have a financial interest in an outcome doesn't mean that their arguments for that outcome are tainted in some undetectable way that can't be revealed by examining the argument itself. The personal stake in the outcome could have been initiated by a belief in the validity of the argument, rather than the opposite.

>It's kind-of like how you rarely trust the implicit or explicit claims in super long infomercials, even without examining the veracity of the evidence for the commercial's claims, or taking the time to think up counter-claims.

You don't need to attack the person making the argument or the medium delivering the argument in order to question its premises.

I think our fundamental disagreement boils down to a question of tractability. It's not that better arguments don't exist. Rather, it's a matter of whether the argument is even worth our time.

> It may be important to note, but it's not argumentation.

Isn't it? The end effect is disregarding a certain set of arguments entirely. If completely ignoring the relevance of an argument isn't a powerful argumentative tool, I'm not sure what is.

> You don't need to attack the person making the argument or the medium delivering the argument in order to question its premises.

But isn't that exactly what's happening in the infomercial case? And surely we shouldn't waste our (bounded, finite) time addressing, point-by-point, infomercial arguments?

I have something more practical and less specific in mind as well. You'll find all manner of conspiracy theorists. The more serious ones can bash you over the head with truth claim after truth claim. At some point, in order to avoid endless rabbit holes, it becomes necessary to disregard certain arguments simply because of their source. I typically do this with a 9/11 truther friend (also chem trains and luminati (sp?) and so on), and only consider his factoids non-fictional when I hear them from other sources as well.

There's also an important distinction between attacking the source per se, and attacking the source's credibility/veracity in the argument at hand. We might be talking past each other due to that distinction as well.

edit: ambiguous grammar

>Isn't it? The end effect is disregarding a certain set of arguments entirely. If completely ignoring the relevance of an argument isn't a powerful argumentative tool, I'm not sure what is.

Why would you do that? For me it's important to note because premises are assertions, and conflicts of interest raise the likelihood that the premises have been deliberately constructed as to be false or misleading. The idea that I would throw out someone's argument because I've decided that it must be wrong in advance seems ludicrous.

Arguments should rise or fall on their own merits.

>But isn't that exactly what's happening in the infomercial case? And surely we shouldn't waste our (bounded, finite) time addressing, point-by-point, infomercial arguments?

Why exactly wouldn't we, and why would we assume that the claims that are made in infomercials are uniformly false? The reason I don't waste my time vetting the arguments in infomercials is because I'm not currently interested in a juicer, not because of a belief that people who make infomercials are liars, no matter how consistent their arguments or verifiable their premises.

>I typically do this with a 9/11 truther friend (also chem trains and luminati (sp?) and so on), and only consider his factoids non-fictional when I hear them from other sources as well.

I don't. The reason I ignore truthers is because their logic is consistently bad, and their premises are usually demonstrably false. Those reasons came after listening to their arguments, not before. There are plenty of questionable elements in the official narrative, too, but they're not in the same order of magnitude as the leaps of logic made by truthers.

> Why would you do that?

Because I have a finite amount of time in the world. Because we need to decide which policy to enact (or not) today and not in 20 years. Because producing and evaluating arguments takes time, and -- most importantly -- because bad actors exist.

> The idea that I would throw out someone's argument because I've decided that it must be wrong in advance seems ludicrous.

Witness e.g. Microsoft trolling standards committees into non-existence back in the day. I'm sure they had an endless supply of valid and brilliantly illustrated technical points to consider, but at some point you have to say "you're a bad actor, and I'm going to start disregarding your arguments because you're making the decision process intractable and, like it or not, something needs to be done eventually".

--

Regarding infomercials:

> The reason I don't waste my time vetting the arguments in infomercials is because I'm not currently interested in a juicer, not because of a belief that people who make infomercials are liars, no matter how consistent their arguments or verifiable their premises.

Right. You have an a priori opinion that you don't need a juicer (or whatever), and you stick with that instead of wasting time considering the commercial's many valid arguments about how buying a juicer will totally enhance your life.

edit: it's not that the infomercial guys are liars per se. It's just that you've decided that evaluating their arguments for why their product will make you happier probably isn't worth your time. And in most cases, I claim that most people make that decision precisely because the source is a paid advertisement. That's an ad hom dismissal of the argument that juicers save marriages (or whatever, can you tell I don't watch infomercials? :-) ).

> Those reasons came after listening to their arguments, not before.

It's the particular person, not any of his theories per se. If one of these theories was mentioned by another person (even someone I didn't know), I would probably investigate if I found the time. But this particular person proved his signal-to-noise ratio is too high long ago, so I stopped listening. That's ad hom, but it's also a cost-effective. I don't want to spend my finite time on this planet objectively analyzing the evidence for every conspiracy theory my friend happens to latch on to. And I think that's rational, even though it's ad hom. And the reason is that evaluating the veracity of arguments is time consuming, and I have finite time.

>Because we need to decide which policy to enact (or not) today and not in 20 years.

It doesn't take 20 years to evaluate an argument. If, on a case by case basis, there's an argument to be made that any of the proposed answers would be better than delaying a decision, then that's an argument too, with premises and reasoning that need to be evaluated.

>Right. You have an a priori opinion that you don't need a juicer (or whatever), and you stick with that instead of wasting time considering the commercial's many valid arguments about how buying a juicer will totally enhance your life.

Maybe I've phrased it wrongly, then. If I'm watching an infomercial, I'm evaluating its arguments. I don't seek out infomercials, because I'm not interested in a juicer. It usually doesn't take very long for infomercials about juicers to make a dubious health claim when I do catch them, though, so in the past I've rejected purchasing one on the merits.

edit: added "usually", in that I've watched infomercials that never made a dubious claim, and I've used products that have been featured in infomercials and been entirely satisfied.

> It doesn't take 20 years to evaluate an argument.

Yes it can, and even in barely contrived scenarios if your proof calculus isn't nice enough.

More importantly, in very practical cases even not assuming adversarial (e.g. Microsoft on standards committees), it most definitely can take 20 years to find a valid argument. Exponential growth is serious stuff. Triple exponential growth is just stupid fast.

If you can't wait 20 years, then you have to move forward with arguments for not-valid-but-sometimes-true cases. And in this case, the agent with the most compute power can generate contradictions for opposing props more quickly and can also generate true instances matching his policy preferences more quickly. Thus, bounded argumentation often inherently favors the more resource-rich adversary.

And the same thing works in more informal settings as well.

One way to cope with this is to disregard arguments from known bad actors. We can discuss rather that's the correct course of action, but the need is a provable fact of the most useful formal systems and an observable fact of informal argumentation (I'm still not sure which you want to discuss, since your assumptions make me think you view all arguments as formal systems, but most arguments -- including this one -- clearly aren't in any meaningful sense of the word formalized).

> And if there aren't better arguments for the other things, then those are the things that should be done.

As a tangential but related point, I think this can be a very dangerous position.

Arguments aren't created in a vacuum, and neither is the lion's share of what makes an argument "good". This is true even discounting the fact that the dominant measure of argumentative quality is largely a function of persuasion.

Producing arguments takes a lot of work and money. That's basically the only reason think tanks and law firms exist. It's also where a lot of marketing money -- political or otherwise -- gets spent. Having really good arguments for something mostly indicates you've invested a lot of time, money, or both into constructing those arguments. (edit: note that this is not an indicator that the thing is actually true, which is the root of the problem).

edit: sometimes, as in science, the motive can be fairly pure. In those cases, even though lots of money is spent constructing the arguments, we can still very much trust the veracity of those arguments.

Sometimes, as in the case of many (not all) think tanks, somehow every argument and study produced always re-entrenches and provides support for a pre-existing viewpoint. I'm therefore far more inclined to view these with suspicion, which is really an ad hom but imho justified.

>I'm not sure what all of this means. The arguments I'm referring to are formal operations on premises. Premises are the environment that arguments live in, and that's not a vacuum. One party claims to be right, states its premises, and how those premises lead to their conclusion. Attacking that argument means pointing out flaws in the reasoning or the premises.

In real life premises are also part of the argument. Aside from mathematics and logic exersizes, there's no discussion that doesn't involve persuassion and arguments about the very premises (and on multiple levels).

That's one of the reasons that the pure mechanistic check of the operations/conclusion part that logical fallacies provide is very nearly useless.

And of course the "checking the premises" part is not mechanistic (that another problem with the formal check), and could also go ad-infinitum. For example, in a moral issue, the premises depend on your values, not some objectively verifiable external truth. On issues of personal interest, the premises depend on what benefits whom. Etc. On political issues the nuances and justifications for the premises are so rich that it's not even worth trying...

Premises are not part of the argument, they're assertions that the argument is based on. The argument is how we get from the premises to the conclusion.

If we don't simply agree on the premises, an argument has to be made to support the ones we don't agree on, and that argument will have further premises. If we don't agree on those premises, an argument has to be made to support the ones we don't agree on, and that argument will have further premises.

Eventually we agree on premises, and work our way back up.

>the pure mechanistic check of the operations/conclusion part that logical fallacies provide is very nearly useless.

I think that it's the only important part of evaluating an argument. So our premises differ. If further arguments that you plan to make rest on the premise that the mechanistic checking of the chain of reasoning from premises to conclusion of an argument is very nearly useless, I may be able to recognize the soundness of this hypothetical argument, but never the validity.

> in a moral issue, the premises depend on your values, not some objectively verifiable external truth.

That I claim to have a moral value is an objectively verifiable external truth.

>On issues of personal interest, the premises depend on what benefits whom. Etc.

When?

>On political issues the nuances and justifications for the premises are so rich that it's not even worth trying...

Isn't the entire political process an act of trying?

> Premises are not part of the argument, they're assertions that the argument is based on

You kind of answer yourself by identifying the fact that premises are often called into question during the course of an argument:

> If we don't simply agree on the premises, an argument has to be made to support the ones we don't agree on, and that argument will have further premises. If we don't agree on those premises, an argument has to be made to support the ones we don't agree on, and that argument will have further premises.

Your parent's argument was that this can go on forever, and never really grounds out, which demonstrates the flaw in the "argumentation is reducible to mechanistic checking" mentality.

I pretty much agree with your parent, because you can see it even when you go all the way to the ground. A lot of evidence cannot get better than statistical, and statisticians sometimes argue about the best way to approach their field. Even for non-statistical arguments, there's some old beef in mathematics about aoc.

But even for most issues where these foundational questions aren't relevant, doing enough work to formalize arguments is intractable.

> That I claim to have a moral value is an objectively verifiable external truth.

That's not very useful when we're trying to make a moral decision and have different competing values. Most political discussions are utilitarian with a mix of various moral imperatives (e.g. optimize well-being, but abide by the constitution's letter and intent).

> Isn't the entire political process an act of trying?

I think he meant it's not worth trying to treat the arguments as chains of reasoning working from a closed set of premises.

Although I agree that the political process in most places could benefit from a bit more logic, it's worth noting that the political process is the act of trying to get power, not trying to reason soundly :-)

>Arguments aren't created in a vacuum, and neither is the lion's share of what makes an argument "good". This is true even discounting the fact that the dominant measure of argumentative quality is largely a function of persuasion.

I'm not sure what all of this means. The arguments I'm referring to are formal operations on premises. Premises are the environment that arguments live in, and that's not a vacuum. One party claims to be right, states its premises, and how those premises lead to their conclusion. Attacking that argument means pointing out flaws in the reasoning or the premises. I'm not talking about convincing people, I'm talking about being right.

I don't understand how the rest of your comment relates to argumentation. It sounds more like a comment on marketing - and while marketing may use different arguments of varying quality, it is not in itself an argument.

> One party claims to be right, states its premises, and how those premises lead to their conclusion.

This is a very vague sentence. What does "how those premises lead to their conclusion" mean?

A formal system?

If not a formal system, then I don't think "being right" is at all well-defined in this sentence. This case converges to the comment below on tractability, which I'll reply to shortly.

If you are referring to a formal system, then the rest of your comment actually is correct. But even in mathematics itself, codifying proof in a formal system is quite rare and practiced only by logicians, certain subfields of CS, and a few early adopters. edit: And, far more to the point, most arguments aren't and can't be mathematical in nature.

> Attacking that argument means pointing out flaws in the reasoning or the premises.

> I don't understand how the rest of your comment relates to argumentation.

The problem is that you're assuming arguments cost nothing -- in terms of money or effort -- to produce. But this simply is not the case. Even in the very best-case scenario of mathematics itself, proof search for even relatively trivial theories is incredibly costly (triple exponential). So even for tiny little factoids that are perfectly well-defined, constructing a valid argument can be intractable.

Now consider the more likely situation where the actual truth of the proposition is effectively unknowable (most of economics and politics) or very difficult to know a priori (a lot of modern science). The resources committed to producing argumentation is probably the most important indicator of the quality of arguments.

> It sounds more like a comment on marketing - and while marketing may use different arguments of varying quality, it is not in itself an argument.

Right, well, I claimed that a lot of marketing is argument construction. So yes, there are other aspects to most marketing as well. But argumentation is certainly part of it.

edit: various small mechanical errors, slight rewording of hastily written sentences.

>If you are referring to a formal system, then the rest of your comment actually is correct. But even in mathematics itself, codifying proof in a formal system is quite rare and practiced only by logicians, certain subfields of CS, and a few early adopters. edit: And, far more to the point, most arguments aren't and can't be mathematical in nature.

I disagree with your premise. I believe that all arguments rest of the degree of truth of their premises, and formal methods are used to prove that the degree of truth in the conclusion springs from the degree of truth of the premises. That is definitional - there's nothing else that I would refer to as an argument. Informal fallacies are also formal fallacies, they're just a way of categorizing fallacies that are written in human language rather than mathematical language.

>The problem is that you're assuming arguments cost nothing -- in terms of money or effort -- to produce.

No, I'm not.

>Right, well, I claimed that a lot of marketing is argument construction. So yes, there are other aspects to most marketing as well. But argumentation is certainly part of it.

I'm only interested within this thread about the portion of marketing that is argument construction and analysis, and not in conflating argumentation with rhetoric, marketing, or propaganda. This is not to deny that those fields do use argumentation, along with the sciences, history, literature, theology, and all other fields.

> I disagree with your premise. I believe that all arguments rest of the degree of truth of their premises

I categorically wasn't referring to the choice of premises, but rather to the cost of finding and evaluating valid arguments themselves, even once premises are settled, for pretty much any interesting/realistic set of premises. That is, in your words, "how those premises lead to their conclusion"

The cost of finding a valid argument, or even deciding rather a prop is true without producing a witnessing argument, is more often than not intractable.

> No, I'm not [assuming arguments cost nothing -- in terms of money or effort -- to produce]

> I'm only interested within this thread about the portion of marketing that is argument construction and analysis, and not in conflating argumentation with rhetoric, marketing, or propaganda.

I assert these two statements are in direct contradiction, since I've provided several examples in this thread of how even and especially appeals to valid argumentation can be used as a form of rhetorical manipulation or propaganda.

Arguments, formal or informal, are costly to produce and even sometimes costly to evaluate.

People with vested interests can and do produce higher-quality arguments than those who don't have the means or simply don't care. Or people who want no action to happen can bog down the conversation. Or people who want something to happen can throw more resources at research and argument construction than their opponents for finite time, and win out.

And all of this even if we assume that every prop has a truth value and that finding that truth value is even decidable, which definitely is not the case.

Reply on news.ycombinator.com