Several things that can all be true at once:
1. Open access publishing is important
2. Peer review is not perfect
3. Community-based vetting of research is key
4. A system for by-passing such vetting muddies the scientific information ecosystem
>>
Yes, this is both a subtweet of arXiv and of every time anyone cites an actually reviewed & published paper by just pointing to its arXiv version, just further lending credibility to all the nonsense that people "publish" on arXiv and then race to read & promote.
>>
Citing a paper that's available through the ACL anthology by pointing to an arXiv version instead is at least the equivalent of putting something recyclable in the landfill, if not equivalent to littering. Small actions that contribute to the degradation of the environment.
Meanwhile, Google Scholar pointing to arXiv versions first is like ... governments providing subsidies to oil companies.
@emilymbender I'm confused, because it seems like the view you're expressing here (re: "3. Community-based vetting of research is key") is implying something like "a small community should be doing evaluation of contributions, and those evaluations should then be deferred to" rather than "members of the small community should find new & effective ways to support a larger and more chaotic community to gain critical evaluation skills". The first feels elitist, and the second feels democratising.
@emilymbender Sorry, I think we must be talking past each other — I absolutely agree that there's value in having work vetted by people who have the expertise to vet them, & I'm not suggesting that arXiv is providing any system of vetting at all!
When I see something that's peer reviewed, I think "someone with status in the relevant field glanced at this, thought it was interesting, didn't notice any egregious errors". That's a higher bar than arXiv, sure, but not by itself "found to be solid".
@galen
First, no it's not elitist to ask for papers to be vetted by people who have the expertise to vet them.
Second, show me how arXiv is a system of vetting? Where do you get to see the results of that vetting, so you can tell which papers have been evaluated and found to be solid?