Hacker News Reader

Comment by zaptheimpaler | Hacker News Reader

If you look up at the sky and say "oh its sunny today", you are overgeneralizing because you haven't seen the rest of the city. If you had worn masks or started being cautious before COVID was widely acknowledged, you were catastrophizing. If you ever talk or act based on a mental model you have of a friend, you are mindreading.

"Cognitive distortions" are the only tools we have to reason about anything in the presence of limited information (which is basically always). Its basically a toolbox to let you discredit any thought whatsoever, which is convenient when a patient writes down negative thoughts and the psychiatrist can just hand them a list. But it would work just as well on positive thoughts or any thought whatsoever.

View on HN

The phrases that the study treats as overgeneralizing are:

all of the time, all of them, all the time, always happens, always like, happens every time, completely, no one ever, nobody ever, every single one of them, every single one of you, I always, you always, he always, she always, they always, I am always, you are always, he is always, she is always, they are always

It's not always overgeneralizing to use one of these. "Every single one of them wore black" is (potentially) completely factual. A large increase in their use in books, however suggests that authors are overgeneralizing more than they did previously.

Phrase list: https://www.pnas.org/content/pnas/suppl/2021/07/22/210206111...

Or that the meaning of these phrases has changed over time.

Word. In fact, most of these phrases, to me at least, imply sarcasm.

To say "social science studies are bullshit" implies a more generalized claim to me than "social science studies are always bullshit"

People use overt generalization words like these, in my experience, to indicate generalization with obvious exception.

This study is selecting for precisely the wrong thing, imo, or is not properly interpreting what they've selected.

I'd like to see their analysis of the phrasing on a forum such as this, where one would have to outline the myriad of corner cases and exceptions lest we get speared by the pedantic knights of the internet.

From the abstract:

> This pattern does not seem to be driven by changes in word meaning, publishing and writing standards, or the Google Books sample

Folks who overgeneralize like that are incredibly easy (and unsatisfactory, due to the predictability of a negative response) to troll. Merely be pedantic, provide a counter example, or suggest a thought experiment in which their generalization may not be true.

Well who is trolling who - if their overgeneralization prompts you to correct them or provide a counterexample, maybe you're already feeding the troll.

Was I trolled into pedantry by your comment?

It’s trolls all the way down. This is why forms of trolling that aren’t blatant flaming get under people’s skin: it’s taking normal good-faith but ignorant user behavior and sending it into the uncanny valley. One troll shutting down a whole community is more often the case of the community having a pre-existing affinity to develop a digital autoimmune disease: once the users start treating good-faith users as potential trolls, the community turns irredeemably toxic.

>>Merely be pedantic, provide a counter example, or suggest a thought experiment in which their generalization may not be true.

It solves one problem but walks right into another. Your counterexample would be observed as so-called sealioning[1], "concern trolling", or one form of the ever-expanding nebulousness of "gaslighting". How does one diffuse obstinate hypocritical crusaders while proving one's own point? The solution doesn't seem to be making a well-made statement. Disagreeing is considered making a "bad-faith" arguments. The answer isn't pointing out their fallacies and thought-terminating clichés. You'll be accused of co-opting an dog-whistling. The only solution I've seen is to do or say as one will as though others don't exist. A lesson learned in one of Aesop's fable, The Miller, His Son and His Ass.

[1]https://wondermark.com/1k62/

I call this talking in superlatives. It probably isn't exactly the same as the article is talking about but it is related. Especially the youth today prefers talking (and thinking?) in black and white. I think it is the effect of two things:

* Media which tries to sell rather mundane news as sensational (exaggerating) * Management literature which tried the last 30 years hard to push people to the limits, to walk the last mile, to strive for the best, greatest, mediocrity is the evil.

This is a common theme in these kind of analyses (even used in economics!) but often the ratio of false positives is sufficiently low for it to not matter enough.

You can find these exaggerations in Shakespeare and Jane Austen.

The authors need to understand the difference between rhetoric and factual reporting.

It's not unusual to say "This is the worst thing in the world" to make a point without believing that it literally is absolutely the worst thing in the entire history of human experience.

Alternatively, none of those bad things are happening to me, so this is literally the worst thing in the world.

My mom does this a lot. Daily, in fact. Has been doing so for decades. She's not self aware enough to do any better. I've tried. She doesn't care. She's not a particularly reasonable, rational person.

Are you overgeneralizing her as someone who is always irrational?

Nope. Unlike you I know her.

No one is particularly rational. Reasonable is maybe 50/50 at best. We can strive though.

As an exception I must certainly agree. People not only aren't being very rational in general, they also really suck at being irrational. Most people are closer to animals than humans.

> "Cognitive distortions" are the only tools we have to reason about anything in the presence of limited information

No, clearly they aren't. But you know what does qualify as a cognitive distortion? That very statement[1]. The "only" way to reason about "anything" based on limited evidence? Really? I mean, Kalman and Bayes would maybe like to have words.

[1] I can see a few, but I'll go with "dichotomous" as the biggest mistake you made. You lept straight from "Sometimes these mental tools produce correct results" (which is true) to "These tools are the only way to produce correct results" (a ridiculous distortion).

One of the results that stood out to me was that (a) dichotomous thinking was the most notable distortion in germany during nazism and (2) they (and the rest of us) are back to peak levels.

I'm talking about reasoning in day-to-day life. It's cool that you are able to use bayesian models & kalman filters to decide whether you should try diet X, or whether someone is lying, or if crypto is a sound investment, or whether to take an umbrella with you when going out today. This is not even scratching the surface of the kinds of nuanced & personal problems that people go to therapists for.

I'm a little dumber, I just overgeneralize to "fad diet X is probably not that great" and skip the hours long scientific paper review & construction of bayesian models and priors.

Here is a recent tweet[1] claiming this paper is badly flawed because of changes in the content of the Google Books data this study was based on. Make sure to update your bayesian models accordingly! Should I lower my prior of "this recent scientific study is trustworthy so i should take it at face value" by 1%, or 5%, or 50%?

[1] https://twitter.com/benmschmidt/status/1419497587296571395

> It's cool that you are able to use bayesian models & kalman filters to decide whether you should try diet X, or whether someone is lying, or if crypto is a sound investment, or whether to take an umbrella with you when going out today.

This part isn't dichotomous, it's just a strawman (something that I'm guessing would go into the "overgeneralizing" bucket in the linked article).

I didn't say you had to use fancy math to live your life. I replied to your statement that one had to use distorted reasoning, and cited the existence of two named techniques (about reasoning from limited evidence) as a fun way of making the point.

My point wasn't about what you said anyway, it was that you were shouting and flaming about it using exactly the techniques the linked article was noting are on the rise. You see that, right?

I wasn't flaming anyone in my original post. Really just expressing some of my skepticisms of cognitive behavioral techniques. Personally I find it difficult to apply in real life when I believe that the technique is strong enough to discredit anything but the strongest scientific claim. The flaming was only to your reply, so you may be indulging in overgeneralization and mind-reading.

It sounds like you are more of a Markov kinda person.

For me, analogies are like an easy crutch that can get you that last mile.

You lept straight from "Sometimes these mental tools produce correct results" (which is true) to "These tools are the only way to produce correct results" (a ridiculous distortion).

Those aren't mutually exclusive. Something can do something only sometimes, and also be the only way to do that thing. If he/she is wrong, It should be easy to prove that their claim is incorrect though by thinking of a counterexample

No, that's not how 'cognitive distortions' work. Think of the process of mental modelling as a hugely complex input-output function. CDS are specific patterns of mental output, that have been associated with changes to the thinking apparatus that we associate with certain psychological disorders.

These are not simple bugs. In fact we know today that working on changing these thinking patterns can in many cases help alleviate the underlying disorders (that's the C in CBT)!

To come back to your example: Being cautious is not catastrophising. Feeling a stress response or even a panic attack or anxiety because of pathogens is (we are taking a broad definition of catastrophising here).

Also beware of the fallacy that positive thoughts are the opposite of negative thoughts - they aren't. They are distinct factors that share some correlation.

> But it would work just as well on positive thoughts

Have you heard of toxic positivity? There’s a growing movement against incessant positivity that has become endemic to a lot of online discourse in particular on Instagram (def not on twitter).

Nothing ever sucks, it’s a challenge! Nobody is sick, they’re a fighter! Nobody is a mean prick, you just need to understand their perspective! Nothing you do is ever stupid, people are just haters!

> Have you heard of toxic positivity?

I suspect it’s a byproduct of “enlightenment now” movement which asserts that humans haven’t had it so good as they do now. Worldwide poverty is the lowest ever, least violence etc. There’s a minor issue of growing inequality and impending climate disasters but that’s all ok and will work out well. Its torchbearer is Steven Pinker and he’s ably supported by, unsurprisingly, the richest 0.1%.

He's also supported by people who believe in empiricism. If it feels like everything is terrible, I suggest watching less news.

You are correct that it is a toolbox. You are also correct that the tools could be used inappropriately. Neither of those truths invalidates the actual tools. Every tool in the world can be used either correctly or incorrectly.

"Oh it's sunny today" doesn't imply anything about the area it is supposedly describing, so it's not an overgeneralization in any way.

"If you had worn masks or started being cautious before COVID" you wouldn't be catastrophizing, because the identifying symbol of catastrophizing is concluding with high confidence a catastrophe from limited or contradictory evidence. We had plenty of evidence for pandemics (and most epidemiologists saw this coming for decades).

I believe you might be confusing cognitive distortions with heuristics (which are similar, but not the same).

Cognitive distortions correspond to irrational thought patterns or lines of reasoning.

If you have limited information, by example, you're not supposed to just jump to a conclusion (except in a matter which requires immediate response), and this is often seen in many mental health disorders where, by example, a person may jump to conclusions such as "everyone hates me" or "I'm screwed for life" from isolated events which predict no such overarching conclusion.

Those kinds of cognitive distortions are not the only tools we have to reason about anything in the presence of limited information, we can often do much better, which is why it's used on CBT.

There’s nuance in almost everything. When writing or speaking you have to generalize, or you’ll just end up rambling.

So when someone says something, you can almost always say they’re being too general and point out some obscure exception. It’s better to just take every statement and implicitly “… in most cases”.

This ignores the whole argument about the proportionality of such rhetorical tools in discourse.

> If you look up at the sky and say "oh its sunny today", you are overgeneralizing because you haven't seen the rest of the city.

Oh? You have seen the sky over the rest of the city. You can see a long way, in the sky.

> You can see a long way, in the sky.

...at least when it's sunny, that is.

Reply on news.ycombinator.com