6 ms·
> "shining a light on it" only causes it to spread more aggressively Do you have any evidence of this? Studies often suggest otherwise, e.g. https://www.scienc
by dbrueck 3y ago
> "shining a light on it" only causes it to spread more aggressively
Do you have any evidence of this? Studies often suggest otherwise, e.g. https://www.sciencedirect.com/science/article/abs/pii/0022103173900723 https://www.sciencedirect.com/science/article/abs/pii/002210...
There are so many problems with censorship, but here are two examples:
1) At a really high level, bad ideas don't hold up well under scrutiny, and censorship literally takes away scrutiny. If you want to weaken the spread of a dumb idea, make its advocates defend it publicly. The most insane and backwards ideas can be made to sound somewhat reasonable in an echo chamber.
2) The quotes around "bad" were a nod to the shifting definition of what is bad. Censorship is appealing when the thing to censor is extreme, but it's a power that is too easily abused. You start off with noble ideas of e.g. censoring "hate" but suddenly realize that that's impossible to define well, and so just about anything can be construed as hateful and, by extension, censorable. Censorship just becomes a cudgel wielded by whoever is currently in power.
- sasmithjr 3y agoI cannot stress enough that I think your second bullet is missing the point by a country mile. The Substack post is specifically talking about Nazi viewpoints, not "bad" viewpoints for some shifting definition of bad. And because we're talking about Nazi views, I think your first bullet then falls apart, too. There's no new scrutiny required for Nazi views because we've already done that work. Additionally, Substack is allowing the monetization of (and thus profiting from) Nazi viewpoints. That's more than "shining a light on it".
- dbrueck 3y agoOk, so play along for a minute: what should Substack's policy about Nazi stuff be, specifically? If you were tasked with defining their policy on just this one subject matter, what would you come up with?
- chris_wot 3y agoDemonetize them. Their only benefit is to publish their work, nothing more. That would not stop freedom of speech, but I can’t understand why you would pay them for their objectionable and vile views.
- dbrueck 3y agoOk, great, so what specifically is the policy you feel that Substack should put in place? I'm inviting you to take a stab at defining at least the rough draft of one. If you can't - if you can't go any further than nebulous terms - then that's just reinforcing my above points and the more general point that it's more or less impossible to do censorship well.
- arcatech 3y ago“We will not host content we don’t agree with on our servers.” Does it have to be more complicated than that?
- dbrueck 3y agoIt depends entirely on your objective. If your goal is to run a site that is your own little corner of the internet where the only views expressed must conform to your precise world view, then that is totally fine and can be very, very effective and (relatively) easy to implement. The content policies for most private companies are basically this, as are many newspapers. Perhaps the most key point is that on those sites, the content is theirs. It all falls apart once you move to user-generated/-contributed content. So, for example, if you want to create a place for the best modern thinkers to come and debate how society should address the modern challenges of our time, how much do you moderate that? Your willingness to interfere with the discussion is directly at odds with your goal of becoming "the" place for this sort of thing. Authors are well aware that the audience/following they build is at least somewhat tied to the platform they use to build it, and many people - possibly some of the people you really want - are going to balk at a policy of "we can take down any post at any time for any reason". It doesn't make your policy right or wrong, it's just at odds with your cool business idea, that's all.