Hacker News Reader

Comment by zeteo | Hacker News Reader

As long as you're not doing pure mathematics, you can't always speak on purely logical ground. Aristotle himself, who invented logic and whose picture is on this poster, wrote a different book called The Art of Rhetoric in which he fully approves of deploying e.g. appeals to emotion or authority. Most of these "fallacies" are actually good heuristics within normal discourse: even if they don't produce full logical confidence, they are still very useful in ambiguous or time-sensitive contexts (a.k.a. everyday life).

View on HN

Yes, this is very true. It is pretty close to impossible to have an argument at bar or internet forum level without committing something that resembles a logical fallacy.

Refer to something written by an expert rather than demonstrating it directly yourself? Appeal to authority.

Suggest that a particular source of information may be unreliaible? Ad hominem

Voice a concern over the direction some policy appears to be taking? Slippery Slope.

Many arguments are also based on statistics rather than directly on logic.

Logical fallacies can be used to expose highly axiomatic thinking though.

Your examples are too broad and/or poorly defined.

An appeal to authority fallacy has to be without reason backing up the appeal other than the fact that they are an authority.

You can suggest that a source of information might be unreliable because of a poor track record in the the domain, or a conflict of interest - that's not ad hominem.

One can speak against the direction of a policy without committing a slippery slope fallacy by not making unsubstantiated claims about what will happen. If you can cite examples of similar policies that have led to effect Y happening with rational explanations for the causes along the way, then that would avoid the fallacy.

Well, for example if somebody asked me the circumference of the earth and I replied with "40,075 km" how do I actually know that?

I certainly have never measured it, so all I can do is cite sources of information which are assumed to be authorities. I can be relatively confident in my reply by citing multiple sources of information but I'm still ultimately relying on something I can't realistically provide other evidence for.

According to wikipedia the second could possibly fall under "Circumstantial Ad hominem" in that suggesting a source has bias does not necessarily make them incorrect. Though this depends on the nature of the argument.

I think the third is more tricky to pin down because it is going to depend again on the nature of the argument. You may suggest concern about some policy without a concrete example of X leading to Y because it may not have happened before. But again this depends on the nature of the argument.

I think my problem is more with the way in which logical fallacies are abused in internet arguments rather than the fallacies themselves. For example somebody saying "You are wrong to be concerned about policy X because slippery slope fallacy".

1. Yeah, this will probably be an accepted fallacy for day to day usage. You could quote leading geologists/physicists and the implication is that they have a vested interest in being correct. But the reasoning there isn't just "Because they're experts, so trust them!", it's "They purport to have measured it and I trust that they did so accurately because of their position as leaders of their field and their desire to remain as such.". Still technically might be a fallacy, but I would accept it in the case of a casual conversation about the Earth's circumference.

2. Agreed that it's contextual. I merely wanted to point out that not all discrediting is ad hominem as you seemed to imply.

3. Also agreed that it's highly dependent on context. Again, just pointing out that one can speak out about policy direction without committing a fallacy.

And yes, I dislike that as well! The fallacy fallacy: https://yourlogicalfallacyis.com/the-fallacy-fallacy

Simple way to verify rough correctness of 24k mile circumference of earth:

Most people know ca is ~3000 miles from ny. Most also know that the time difference is 3 hours. So 1 time zone is approx 1000 miles. Circumference = 1000 mi x number of time zones.

If you've got nothing to go on but second-hand information that you can't validate yourself, then it's perfectly reasonable to use qualities like "credibility" and "authority" as a sort of Bayesian indicator of the likelihood that the statement is a valid one.

The "appeal to authority" is only a practical fallacy in a situation where it is feasible to directly evaluate the claim against reality itself.

> Refer to something written by an expert rather than demonstrating it directly yourself? Appeal to authority.

I'd say that for practical matters, there are degrees to this. Example: Health. There's a difference between referring to a doctor versus a Hollywood celebrity.

> Voice a concern over the direction some policy appears to be taking? Slippery Slope.

Most of the time I see people calling out a "slippery slope fallacy", they're often applying it erroneously. The actual fallacy is in asserting that a step in a particular direction logically entails a further step in that direction; it is fallacious in most situations to assert the slippery slope as a deterministic necessity.

But that has no bearing on whether or not the slippery slope actually exists; just because it isn't logically entailed doesn't mean that, in the empirical context of the matter, a step in one direction doesn't actually make a further step in that direction more likely.

There is an empirical slippery-slope phenomenon in political debate, related to the concept of the "Overton Window" [1], and it exists contingently irrespective of whether or not it's logically necesssary.

[1] http://en.wikipedia.org/wiki/Overton_window

The problem with most of these (informal) fallacies is that they are attempts to prevent argument, either by shutting the other party up, or by diverting the discussion away from the matter at hand. When they are used like that they could be considered incompatible with a commitment to reason, or as fallacies.

The philosopher Arthur Schopenhauer documented these tactics humorously in The Art of Controversy (Chapter III):

http://ebooks.adelaide.edu.au/s/schopenhauer/arthur/controve...

One of my favorite books is Thank You For Arguing, and it's a bit more accessible than an Aristotle translation, for those who want a faster introduction to why fallacies are okay.

http://www.amazon.com/Thank-You-Arguing-Aristotle-Persuasion...

There are two main reasons that fallacies are bad. The first is that they're not rigorous, which is why parent mentions pure mathematics. Perfect logic matters, sometimes. Strong philosophical positions are like this, too: the kind that takes you years to develop and stabilize. But it's actually pretty rare for anyone to spend much time here.

The second is that fallacies are meant to contort the mind. Which isn't a bad thing, but if you're not equally equipped in logic and an understanding of fallacies yourself, deft usage of fallacy will run circles around you. While virtually all thought experiments are fallacious by nature, if you can't figure out where the boundary between allegory and reality are, you're going to have a problem.

Yeah, a key feature of Greek and Roman education was focusing on winning the argument on both a logical and emotional front--they taught that if you didn't use both effectively you weren't really doing a good job.

And much debate has nothing to do with deducing conclusions from an agreed-upon set of hypotheses at all. A lot of it is about offering different sets of hypotheses and evaluative criteria, and attempting to convince others to adopt the starting point of your reasoning, not its final conclusion.

There's also a lot of useful metadata that gets communicated via "fallacious" argumentation: the intentions and credibility of the speaker, alignment of incentives, presence of alterior motives, etc. This is all entirely appropriate, especially - as in politics - when the debate is about an ought rather than an is.

You are correct. It's not because you use a fallacy in your discourse that you are wrong.

"The fallacy fallacy: You presumed that because a claim has been poorly argued, or a fallacy has been made, that the claim itself must be wrong."

Reply on news.ycombinator.com