zlacker

[parent] [thread] 15 comments
1. dang+(OP)[view] [source] 2022-07-14 23:59:06
On HN we try to go by article quality, not site quality: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so.... The author names here are probably more important than whoever owns the site. Their credentials are on the line if the sources they claim to be quoting are made up. That doesn't prove anything, of course, but one's always guessing with these things, and readers need to make up their own minds.

I understand the argument you're making, and it's not an obvious call, but I think it comes with more downside than upside, at least for HN. It's a trope of tribal internet argument (I mean in general—not talking about you here) to follow a "DAG of shame" in which you hop from any node to the most shameful associated thing, with the intention of discrediting the node from which you started. The problem is that each of those hops loses a lot of information, and one ends up in places that aren't particularly relevant, like whatever that university project is.

What's bad about this for HN is that it makes threads more generic, predictable, and repetitive. It also polarizes discussion along the most intense axis. All of this makes discussion less interesting and more inflammatory.

So while it's not an obvious call (more like 60-40 than 80-20) I think we're better off as a community to resist the habit of replacing topic X with the biggest or most shameful other-topic-Y that the dots connect to. It's not that there's no value in it, but it's the wrong move, given what we're trying to optimize for (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...).

replies(6): >>KerrAv+f1 >>pvg+n1 >>gpt5+A1 >>dredmo+v2 >>mmmpet+0t >>tptace+pu
2. KerrAv+f1[view] [source] 2022-07-15 00:08:45
>>dang+(OP)
I understand this argument, and for technical matters, that's fine, but I don't think this is sensible as a general guide for medical or political sites like this one. These are very common snake-oil pits, and it's difficult for people without the necessary specialization of knowledge to discover the pitfalls. The source is sometimes the issue!
replies(1): >>dang+22
3. pvg+n1[view] [source] 2022-07-15 00:10:21
>>dang+(OP)
Seems like a fairly straightforward call by HN's own practices - if these big-if-true claims are supportable, there's going to be more neutrally sourced, reputable reporting on it soon enough. So it's ok to wait. This wouldn't generally apply to a site that happens to have an ideological or political slant but seems ezmode for regular news, especially public-health news.
4. gpt5+A1[view] [source] 2022-07-15 00:11:17
>>dang+(OP)
Thanks for fighting against ad hominem rhetorics
replies(1): >>dredmo+n3
◧◩
5. dang+22[view] [source] [discussion] 2022-07-15 00:15:26
>>KerrAv+f1
For better or worse, the principle here has always been to trust readers to be smart enough to make up their own minds (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...). I'm not saying we all are (nor that I am), but I think it's the right principle, especially for this place, where trying to tell users what/how to think is guaranteed to provoke a backlash (quite separately from political position). The consequences of dropping that principle seem pretty negative to me.

Following standard ideological grooves to discredit the other tribe's sources is not acting from specialist knowledge in any case. The only specialty at work in such discussion is the specialty of internet battle. That's ultimately just a way to turn every thread into a boring, if intense, flamewar.

One of the things that follows from HN's core principle of intellectual curiosity (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...) is the principle of diffs (https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...). Diffs are what's interesting. This is the positive formulation of the principle that repetition is bad for curiosity (https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...).

For topics that burn hot, like political/ideological ones, this "diff" principle implies that discussion is most interesting (and therefore best for HN) when people don't take the obvious path from their own initial position—that is, when they don't repeat the reaction that they've had most often before. That is a hard thing to ask on the hottest topics, which tend to melt into a few (well, two) monolithic piles of tar. But I think it follows from the principle.

Here's another thing that I think follows, and is even harder to swallow. To the extent that someone has strong political/ideological views, if they're not seeing articles on HN that they strongly disagree with, at least semi-regularly, then there's probably something wrong with HN*.

That isn't always great for community spirit because it only takes a few disagreeable data points before the mind starts to defend itself with a "this place sucks" reaction (and there are people on all ideological sides who develop such reactions). I wrote about this here, if anyone wants to read more: https://news.ycombinator.com/item?id=23308098.

* To ward off one common misunderstanding: that is not a Goldilocks argument for split-the-difference centrism! It's an argument for unpredictability. Since centrism is just as predictable as other ideologies, it should encounter just as much to be put off by.

replies(2): >>ianai+83 >>jonste+lp
6. dredmo+v2[view] [source] 2022-07-15 00:20:06
>>dang+(OP)
HN also recognises and deprecates numerous sources which have been shown to place little regard on truth or to generate needlessly and gratuitously provocative articles.

It's possible to give fair hearing without opening the door to bullshit once the bullshitter is known and demonstrated. Editorial voice is significant, as HN's own moderation strives toward and demonstrates.

◧◩◪
7. ianai+83[view] [source] [discussion] 2022-07-15 00:25:03
>>dang+22
Aren’t some topics just out of bounds for discussion here though? Topics by their nature which produce discussion unsuitable for what you want at HN. I’ve seen commenters push back on music threads, for instance. And much sex talk seems to quickly devolve from anything productive here.
replies(1): >>dang+v4
◧◩
8. dredmo+n3[view] [source] [discussion] 2022-07-15 00:26:32
>>gpt5+A1
Ad hom is use of an irrelevant characteristic to impeach a witness or malign a viewpoint.

"You wear funny hats so you're not a credible expert on waffle irons" would be an ad hominem.[1]

Noting the reputation and past history of a speaker or source, or noting well-known rhetorical red flags (http://classics.mit.edu/Aristotle/sophist_refut.1.1.html) is what we'd now call a strong Bayesian prior. It's not a proof that a source, speaker, document, or publication is wrong, but it's a fairly strong grounds for suspecting that might be the case.

https://www.thoughtco.com/ad-hominem-fallacy-1689062

And in a world in which asymmetric costs favour bullshit, it's a useful and often necessary approach.

https://en.wikipedia.org/wiki/Brandolini's_law

________________________________

Notes:

1. In practice, "funny hats" might well be references to non-Western clothing (see The Little Prince for a fictional illustration), or speech, or gender, or religous affiliation (religious exclusion was common in top US universities well into the 1970s). It's still practiced in many regards. What this ignores is the specific capabilities or validity of claims or methods.

replies(1): >>bmelto+vl
◧◩◪◨
9. dang+v4[view] [source] [discussion] 2022-07-15 00:35:48
>>ianai+83
> Aren’t some topics just out of bounds for discussion here though?

No, if you look in the site guidelines you'll see that they say "Most" stories on certain themes are off topic, and that word is there intentionally. It allows for exceptions, especially when there's either something interestingly different about a story, or some significant new information to discuss.

https://news.ycombinator.com/newsguidelines.html

For past explanations about how we think about this in terms of political topics, see https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so.... There are lots of links there to where I've explained this in detail. If anyone has a question that I haven't answered there, I'd be happy to take a crack at it.

Stories drawn from the arts, like music and also literature and painting, or more generally from history, archaeology, you name it, are most welcome here as long as they offer something of intellectual curiosity. So if HN commenters are pushing back on a story just because it's about e.g. music, that would be bad. (But if it were a gossipy story about a famous musician, say, that would be different.)

Sex is its own special case in all things, so we would have to talk about that separately.

replies(1): >>ianai+Bm
◧◩◪
10. bmelto+vl[view] [source] [discussion] 2022-07-15 03:04:56
>>dredmo+n3
That is not a correct use of ad hominem. Translated literally, it is "against the person," which is what an ad hominem attack is -- an argument against the bearer of the idea that does not argue against the idea itself.

An argument against the idea was almost presented (anonymous staffers whose quotes cannot be independently verified) but was let down by its conclusion (it falls to the reputation.)

There was nothing in OP's argument that contradicted the claims that were made other than those that go against the reputation of the author, ergo even if they are valid, they are definitely _ad hominem_.

replies(1): >>tptace+Ru
◧◩◪◨⬒
11. ianai+Bm[view] [source] [discussion] 2022-07-15 03:18:32
>>dang+v4
Good to know. Thank you.
◧◩◪
12. jonste+lp[view] [source] [discussion] 2022-07-15 03:48:14
>>dang+22
> the principle here has always been to trust readers to be smart enough to make up their own minds

That's a fine principle. By extension, it should also be fine for commenters to note facts about sources. You are right that internet discussion can be derailed by the DAG of association, but internet discussion based solely on "what does the article say?" is naïve, amounting to borderline sealioning.

Part of critical thinking and reading is understanding the POV of the author(s) and publisher(s), and considering their motivations and incentives.

replies(1): >>dang+S72
13. mmmpet+0t[view] [source] 2022-07-15 04:34:23
>>dang+(OP)
If the entire article is based on "This guy told me this off the record" then what do we have to go by other than the authority of the source? "Article quality" is completely based on verifiability of evidence here. I actually somewhat agree with the main thrust that CDC is completely politicized and can no longer be trusted. This article does not help and is just noise.
14. tptace+pu[view] [source] 2022-07-15 04:52:28
>>dang+(OP)
I don't read the parent comment as calling for Bari Weiss's substack to be banned from the site, but rather just providing some context about what the site is. I'd have maybe pointed out more about the contrarian impulse of the site, and how that interacts with this topic (being contrarian about public health policy surrounding COVID is a click magnet), but either way it seems like a colorably valuable comment.

If the comment opened with something like "I stopped reading at commonsense.news because Bari Weiss is bad", that'd be a different matter, but they have a substantive critique of the actual article.

Given the weak sourcing, it feels like this article, in particular, flunks the "divisive subjects require more thought and substance" test.

I loathe Bari Weiss, so, grain of salt on all this.

◧◩◪◨
15. tptace+Ru[view] [source] [discussion] 2022-07-15 04:59:47
>>bmelto+vl
This would be a more interesting rebuttal if the story itself didn't stake the validity of these arguments on the identities of the people making them. Given that it's based in part on an appeal to the authority of, e.g., anonymous NIH scientists (in which group? with what kind of tenure?), pointing out the anonymity of the sources is not at all fallacious.
◧◩◪◨
16. dang+S72[view] [source] [discussion] 2022-07-15 18:00:46
>>jonste+lp
Sure, but "noting facts about sources" is so broad that it can cover everything from adding neutral, interesting information all the way to tribal warfare. It's the latter that I'm objecting to (for HN), for reasons I think I've described already: it makes threads more repetitive, predictable, and nasty.

I don't think the DAG-of-shame game really has to do with curiosity about facts. It has to do with tarring ideological enemies. Maybe they deserve it, maybe they don't, but it's not the quality of discussion we want here.

Rather than an ambiguous phrase like "noting facts about sources", I think we're better off applying the clearer distinction between curious conversation and ideological battle. There's a binary distinction between those two things (as binary as these things can get), and we know which one we want on this site.

Moreover, one destroys the other, so it needs to be actively moderated. I don't just mean what moderators do, but the general sense of dampening excesses and avoiding extremes. We want a culture of moderation on HN—not for its own sake, but because curiosity only flowers in a temperate climate.

(By the way, I'm not disagreeing ideologically with any of the comments that I'm objecting to here. This is about discussion quality and attempting to organize the site around one specific value: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor....)

[go to top]