[haswell]

> I also notice that argument by analogy is being over-used

This one seems especially pernicious, not because of extremely over-wrought comparisons, but because sometimes the analogy fits really well on the surface. But beyond the structural fit, it does not really help prove anything.

Too often I'll encounter an analogy wielded as if it proves the underlying point, when the reality is that it breaks down quickly if you dive into the details.

Analogies can be great to help establish new mental models, or to try out an idea with terminology that people already understand, but can be quite misleading. Better used for learning than trying to prove things.

[ben_w]

Analogies are like a box of chocolates: overused clichés, but we hand them out and gobble them down with delight.

[efitz]

Analogies are decorative writings, not arguments.

[Jensson]

They can be arguments, for example if you want to show that the other person has inconsistent views due to feelings or bias an analogy between the two inconsistent scenarios can make that clearer. I've changed my mind thanks to analogies many times.

[haswell]

To split hairs a bit, analogies can support arguments, but are not considered a conclusive form of evidence/argumentation in and of themselves in formal logic.

While the analogy may have been instrumental in helping change your mind, it likely did so by helping you understand the actual underlying argument.

[Jensson]

If a human based their views solely on formal logic then they wouldn't have an inconsistent world view in the first place, so I don't see why you bring up formal logic here.

> While the analogy may have been instrumental in helping change your mind, it likely did so by helping you understand the actual underlying argument.

The world would be a much better place if that was true, but sadly people base their world view largely on feelings and those feelings often doesn't care about the underlying arguments but they can feel the analogies. That goes for you and me as well, feelings are a fundamental part of human thinking, you can't just ignore that just because formal logic says it isn't important.

For example, a person might say that they are against racism but they are pro discriminating against white people. That aligns with their feelings, but it is inconsistent and you would need something more than formal logic to make them see that inconsistency. And once they see it you didn't do it by making them understand a formal argument, you did it by changing how they feel about things, they already agreed with you that discrimination is bad they just didn't apply that consistently due to their feelings clouding their minds.

[haswell]

I agree that people are often swayed by feelings, and that we don’t operate as perfectly logical/rational beings.

But if the only thing that convinced you of a thing is an analogy or emotional appeal, and if in being convinced you learned nothing about the underlying argument the analogy is supporting, you are susceptible to being convinced by similarly compelling analogies or emotional appeals that may or may not have any grounding in a solid argument.

I’m not claiming that people can’t be convinced this way. I’m pointing out that this form of persuasion is problematic and insufficient. “Argument from analogy” is considered a fallacy for this reason. It can lead people to take on new beliefs for bad reasons, even if the position they take is the “right” one.

While emotions absolutely influence our beliefs, it’s not accurate to downplay the role of formal logic, which is often implicitly invoked by emotional dialogue. The two are not mutually exclusive, and they work together.

For example: in a discussion about climate change, a purely logical presentation of facts about the temperature of the ocean and receding ice is not compelling without understanding the implications. Painting a picture of potentially cataclysmic outcomes and mass extinction events and migrations evokes an emotional response that is also necessary for humans to take action.

In this example, either one without the other can be problematic. Pure facts logically presented are hard to interpret, especially if you aren’t a climatologist. And if there is no logical foundation whatsoever, the argument is on shaky ground and the person who now believes it will have no reason not to believe the next emotionally compelling thing.

[defrost]

> If a human based their views solely on formal logic then they wouldn't have an inconsistent world view in the first place,

Can you remind the class about Gödel's second incompleteness theorem and what it says about consistency in formal logic systems?

[ggm]

I love the orange cream. I hate the nougat.

[dylan604]

what if you're wrong

[ggm]

Then I have to change my mind. But I think right now, it's not that I'm wrong to prefer orange cream, it's that I can't justify it from objective criteria except that one time I split a tooth on hard nougat which might colour my preferences

[dylan604]

If you're not wrong, then I'm wrong, and well, that's just not possible ;^)

[Terr_]

Analogy discussions break down when people aren't in agreement about which features are or aren't important to map across... or worse, when one or both sides haven't even considered what they are.

[namaria]

Someone making a point through analogy is pretty much a caricature of bad reasoning. It's a pretty common trope.

[ruszki]

So far I haven’t seen an argument in this thread for why it’s bad reasoning when you want to show that the reasons are not what’s stated. It was just stated that it’s bad. So why is that?

[namaria]

I've honestly tried but I can't make out what you mean here:

> you want to show that the reasons are not what’s stated

[ruszki]

Somebody states that they think X because of Y, and they don’t say anything else (Y can be also a group of reasons). You show an analogy where Y would cause Z too (provided the logic is solid). The other party states that they think differently about Z - the reason doesn’t matter.

In this case, either Y isn’t important at all, or there is also something else besides Y, which is not stated. In other words, X is not because of Y logically.

[namaria]

I don't see a problem or a point. A cause can have more then one effect, a logically sound analogy doesn't make all analogies logically sound. And producing causal arguments is inherently hard, we've spent centuries holding irrelevant things as fundamental causes of phenomena, and we still don't know fundamental causes for most observations we have produced.