Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Any claim without a URI should be treated as suspicious

shkspr.mobi

Post reply on HN
Print view View original

141–150 of 282 posts

Re: Any claim without a URI should be treated as suspicious

#141

Post by mistermann »

zentiggr wrote: post #80
082349872349872 wrote: post #74

Earlier quoted context omitted.

Page 58 of "The Art of Conversing With Men" (an 1805 translation of a 1788 work by Knigge) shows that trolls were a thing in the eighteenth century: > "There are people who pretend to know every thing better than others, contradict every one, frequently against their own conviction, merely for the sake of disputing. There are others who are fond of speaking in parodoxes , and accustomed to maintain assertions which n…

It is a bit of an art to separate those who are discussing a point for sake of clarity or persuasion, from these bastards whose only goal is to get a rise out of you or consume your time with their own amusement. Disengaging from those conversations is easier though. Don't answer, walk away, just stop playing their game. Although there is a part of me that wants to shout them down and ridicule them, that plays right…

> Disengaging from those conversations is easier though.

Assuming you've accurately identified one.

> Don't answer, walk away, just stop playing their game. Although there is a part of me that wants to shout them down and ridicule them, that plays right into their bs as well.

Assuming that you have not misunderstood their intended message, or made a mistake in the formation of your belief on the matter.

My comment right here could easily be considered (evaluated as) an example of "BS". However, here it is useful to realize that there are many ways to evaluate many things, that can produce extremely different conclusions. Sometimes conclusions can even be completely contradictory, while simultaneously being objectively correct.

Re: Any claim without a URI should be treated as suspicious

#142

Post by staycoolboy »

How are citations not the fallacy of "appeal to authority"?

Citations are basically an attempt at X.509 or Certificate Authority "root of trust": follow the certs until you find one we all agree is "trustyworthy".

Except in this case, that "someone" may or may not be provable, because the root of trust could potentially go back centuries.

I'm not impugning the work of historians, I realize this is part of what they do: follow the citations, cross-correlate, and seek "the truth" through painstaking research.

But outside of extensive historical effort, it seems like a pretty flimsy strategy for contentious topics.

Re: Any claim without a URI should be treated as suspicious

#143

Post by c3534l »

btrettel wrote: post #47

As many of those who have read academic papers know, even claims with citations are often suspicious. Time and time again I've found cited sources that don't say what was claimed. Sometimes, citations are not specific enough (e.g., citing a huge book rather than the specific page with the evidence). So, better than giving a URL would be to give a URL with a "fragment" if possible, i.e., #id. https://en.wikipedia.org/…

I remember at a certain point I realized that the citations on cracked.com were often abused or flat out debunked the very thing the author claimed they did. I hadn't given up on the site yet, so I would post in the comments what the citations actually said. I'm still baffled by the generally negative reactions to doing so, but I do remember people saying things like "why should I trust you, some random dude in the comments, rather than a professional writer whose job it is to research these things?" Because it's a link to a website that anyone can read! But people will argue with you for 5 minutes rather than click on a link and read the surrounding few paragraphs to see that the writer took the source out of context.

Re: Any claim without a URI should be treated as suspicious

#144

Post by Goronmon »

CM30 wrote: post #117
btrettel wrote: post #47

As many of those who have read academic papers know, even claims with citations are often suspicious. Time and time again I've found cited sources that don't say what was claimed. Sometimes, citations are not specific enough (e.g., citing a huge book rather than the specific page with the evidence). So, better than giving a URL would be to give a URL with a "fragment" if possible, i.e., #id. https://en.wikipedia.org/…

I still remember one fake news site that took this to the extreme, in that it linked to a story that flat out didn't exist on the site it linked to. You clicked on the link, and you just ended up with a 404 error page. The fact that fooled anyone should probably give you a worrying insight into how many people are willing to trust/distrust links without even clicking on them. But yeah, either way, make sure the link…

I still remember one fake news site that took this to the extreme, in that it linked to a story that flat out didn't exist on the site it linked to.

More common what I see is these sites just link to other articles on the site as a "source" for claims made in more recent articles. You almost never see sources to external sites or resources, its just internal site links all the way down.

Re: Any claim without a URI should be treated as suspicious

#145

Post by mistermann »

donarb wrote: post #90
softsound wrote: post #81

Earlier quoted context omitted.

Now if only my family could believe Snopes. They are too set in their thoughts to believe not everything is some conspiracy theory. Now when I debunk stuff off Facebook they continue to disagree with me because I don't have "all the facts", not like they do either. Sometimes a simple thing is just too simple for them, those coin shortages really are some sort of mass government takeover and not just caused by people…

They need to understand that Snopes is not an authoritative source on its own. It provides links to other sources that provide information about the claim.

People also need to understand that those links are also not authoritative, and that what links are provided (and ones that are not) are a result of human choice, and humans are known to be fallible in many different ways.

Re: Any claim without a URI should be treated as suspicious

#146

Post by koheripbal »

btrettel wrote: post #47

As many of those who have read academic papers know, even claims with citations are often suspicious. Time and time again I've found cited sources that don't say what was claimed. Sometimes, citations are not specific enough (e.g., citing a huge book rather than the specific page with the evidence). So, better than giving a URL would be to give a URL with a "fragment" if possible, i.e., #id. https://en.wikipedia.org/…

What might be better would be some sort of reputation system. Something where I can see that other people that have looked into the person's citations confirm they actually say what they claim they say, on average (regardless of agreeing with their conclusions).

Reputation is everything, and it's so incredibly scarce today.

We could have different dimensions of Reputation (citation validity, logical argument validity, unsupported claim predictiveness, etc...)

Re: Any claim without a URI should be treated as suspicious

#147

Post by dqpb »

koheripbal wrote: post #146
btrettel wrote: post #47

As many of those who have read academic papers know, even claims with citations are often suspicious. Time and time again I've found cited sources that don't say what was claimed. Sometimes, citations are not specific enough (e.g., citing a huge book rather than the specific page with the evidence). So, better than giving a URL would be to give a URL with a "fragment" if possible, i.e., #id. https://en.wikipedia.org/…

What might be better would be some sort of reputation system. Something where I can see that other people that have looked into the person's citations confirm they actually say what they claim they say, on average (regardless of agreeing with their conclusions). Reputation is everything, and it's so incredibly scarce today. We could have different dimensions of Reputation (citation validity, logical argument validity…

Do you think reputation can be manipulated?

Re: Any claim without a URI should be treated as suspicious

#148

Post by herbstein »

AnHonestComment wrote: post #87
SkyBelow wrote: post #78

Earlier quoted context omitted.

>I largely given up providing links to information when called out as well. It's not worth the time. Demanding citations you don't plan to read, to either waste a person's time or to make them look unreliable, is a tactic I've seen used so often I normally don't care to provide links. Instead, I'll just point out some 'so you agree with my argument but doubt data backs it up?' at which point, if they respond, it'll b…

As someone who gets censored for wrongthink routinely, I’ve taken up having GitHub repos of links so it’s easy to support my point. Case in point: I’m “banned” on HN because @dang thinks my genuine view, calmly stated is inherently a flame war and I shouldn’t be allowed to join the conversation.

I don't know which view it is you're talking about, but what @dang did isn't universally wrong. If the view is something that is as close as possible to objectively wrong, in the moral sense, as something could be, I'd say it's justified.

An example of this could be genuinely stating "Hitler did nothing wrong" or "black people deserve slavery". You know, reprehensible stuff. There's no reasonable way to have that discussion. How do you debate whether a large minority is inherently sub-human? Like, I'm sure a Neo-Nazi would be able to calmly explain their views without attacking me (a white man) personally, verbally, or physically. But their view is inherently an attack on the humanity of other people.

To be clear, I'm not saying you are actually a Neo-Nazi or a racist. They're just very, very clear-cut examples of the point I'm trying to make. Another example could be that someone stating that every LISP programmer is a child molester. There's no flamewar in the comment itself, but it's a view that can't be expressed without inciting one.

Yet another example could be thinking that every person earning more than $50k is a capitalist pig and deserve the guillotine. Again, calmly stated as a matter of fact. But it's also calling for the execution of a large part of this site's audience. It would inherently cause a flamewar too.

All of these examples are extreme, but I think the point that some views inherently having a piggybacking flamewar has been made. Where @dang decides to draw the line between controversial point with legitimate room for discussion, and views so abhorrent or nonsensical that it's easier to simply ban users perpetuating them is up to their discretion.

I also don't think @dang bans people because of a single comment, but rather someone continually pushing the view that inherently leads to flamewars.

Re: Any claim without a URI should be treated as suspicious

#149

Post by AnimalMuppet »

btrettel wrote: post #47

As many of those who have read academic papers know, even claims with citations are often suspicious. Time and time again I've found cited sources that don't say what was claimed. Sometimes, citations are not specific enough (e.g., citing a huge book rather than the specific page with the evidence). So, better than giving a URL would be to give a URL with a "fragment" if possible, i.e., #id. https://en.wikipedia.org/…

> Time and time again I've found cited sources that don't say what was claimed.

I've seen that here on HN, too. Person 1 claims X. Person 2 asks for evidence. Person 1 posts a URL, as if it's evidence for X. But when you read the contents carefully, you realize that it doesn't actually provide any evidence for X. It provides evidence for something that's somewhere in the same general topic as X, but that's not what Person 2 asked for.

Re: Any claim without a URI should be treated as suspicious

#150

Post by AnimalMuppet »

clairity wrote: post #139
btrettel wrote: post #47

As many of those who have read academic papers know, even claims with citations are often suspicious. Time and time again I've found cited sources that don't say what was claimed. Sometimes, citations are not specific enough (e.g., citing a huge book rather than the specific page with the evidence). So, better than giving a URL would be to give a URL with a "fragment" if possible, i.e., #id. https://en.wikipedia.org/…

you should be able to make an argument without resting primarily and directly on other people's words, even if you largely rest on other people's ideas. whether it convinces any particular person is largely out of your control, and shouldn't consume the debate. citations don't change that calculus.

Citations can give data, though. If your ideas are based on more than just your own experience, you need external data.
Post reply on HN