4 ms·

The institution doesn't get that money: literally not a single dime of it goes to the authors, their departments, or their institutions. It goes to publishers.

The institution doesn't get that money: literally not a single dime of it goes to the authors, their departments, or their institutions. It goes to publishers. You don't need to be friends-of-friends either. Search for the authors and if the PDF is not on their webpages, shoot one of them an email. All it has to say is "I'm interested in your paper [title], but don't have access to [journal]. Could you send me a reprint/preprint? Thanks." It makes my day when people do this. Getting a paper published is a huge slog with overwhelmingly negative feedback, so it feels great to know that someone actually cares about what you wrote.

lucb1e 5y ago
> It makes my day when people do this. Getting a paper published is a huge slog with overwhelmingly negative feedback, so it feels great to know that someone actually cares about what you wrote. Wait, I don't get this. Probably this is very dumb, pardon my nonacademianess. My first thought is: then why don't you just host it on your own website? Then you can see exactly how many people download it. Heck, it could even be in an accessible HTML format (I'd love this instead of basically-image formats like PDF) and you can check if people leave the page after 5 seconds or if there are 30 minutes worth of mouse movement events. Not to say that you need to track this all identifiably, but the server could ingest "user still active on page" events and aggregate that per IP address or something until the session goes cold, then convert it to a mere "visitor #9001, stayed 27 minutes". I'm probably already overthinking this because I've never seen anyone host their own paper in the first place, nor any other format than PDF. Why doesn't anyone do that? Sending an email to request the paper is much slower than just having access (be that through sci-hub, a friend that you can see is online and whose university still supports the publisher system, etc.), so I don't think that's a good solution even if I understand perfectly why you love to hear of your readers. Imagine you had to email the authors of every link here on HN every time you think a headline sounds interesting; it's just not a solution to paywalls with ridiculous costs per article.
- AlbertCory 5y ago
  Maybe you've heard of copyright and contractual obligations?
- mattkrause 5y ago
  This is a weird edge case. - For most academic articles, the audience is almost entirely other academics with institutional access. If so, getting a paywalled article usually just requires signing into the VPN and reloading the page. If your university has a subscription--and it likely does to "core" journals, you're good to go. If not... - Funders have been pushing for "open access" to research findings. Work funded by the US government needs to be freely available online within a year of publication. You should therefore either be able to access the article text somehow, either through the journal itself, a preprint server (arXiv or one of its offshoots), or a repository (PubMed Central, DTIC, etc). - On top of that, most labs do have webpages that contain their publications. Google scholar is pretty good at finding them too. This covers ~99.9% of cases, and they're all instant once you know how to use them. My point in the original comment was if that these fail, you should shoot the authors an email instead of paying. A few publishers do hold a fairly hard line against publicly posting an article, but even they typically allow you to provide a copy "on request". Also, in case it's not clear, academic publishing is only partly about actually disseminating information. A huge part is getting the "imprimatur" of peer review generally, and from certain journals specifically. A brilliant self-hosted document might impress people within a very specific subfield that are motivated and able to evaluate the work. However, its value will be unclear to others who fund/hire/evaluate you. The Dean, for example, likely has neither has the background nor the time to decide if your technical blog merits tenure. He/she does know that Nature is very selective, publishing there means that your work is held in high regard by your field, and therefore you deserve tenure. This holds true for funders too. It's not a great system in many regards, but it is how things currently work.

lucb1e 5y ago

> It makes my day when people do this. Getting a paper published is a huge slog with overwhelmingly negative feedback, so it feels great to know that someone actually cares about what you wrote. Wait, I don't get this. Probably this is very dumb, pardon my nonacademianess. My first thought is: then why don't you just host it on your own website? Then you can see exactly how many people download it. Heck, it could even be in an accessible HTML format (I'd love this instead of basically-image formats like PDF) and you can check if people leave the page after 5 seconds or if there are 30 minutes worth of mouse movement events. Not to say that you need to track this all identifiably, but the server could ingest "user still active on page" events and aggregate that per IP address or something until the session goes cold, then convert it to a mere "visitor #9001, stayed 27 minutes". I'm probably already overthinking this because I've never seen anyone host their own paper in the first place, nor any other format than PDF. Why doesn't anyone do that? Sending an email to request the paper is much slower than just having access (be that through sci-hub, a friend that you can see is online and whose university still supports the publisher system, etc.), so I don't think that's a good solution even if I understand perfectly why you love to hear of your readers. Imagine you had to email the authors of every link here on HN every time you think a headline sounds interesting; it's just not a solution to paywalls with ridiculous costs per article.

AlbertCory 5y ago
Maybe you've heard of copyright and contractual obligations?
mattkrause 5y ago
This is a weird edge case. - For most academic articles, the audience is almost entirely other academics with institutional access. If so, getting a paywalled article usually just requires signing into the VPN and reloading the page. If your university has a subscription--and it likely does to "core" journals, you're good to go. If not... - Funders have been pushing for "open access" to research findings. Work funded by the US government needs to be freely available online within a year of publication. You should therefore either be able to access the article text somehow, either through the journal itself, a preprint server (arXiv or one of its offshoots), or a repository (PubMed Central, DTIC, etc). - On top of that, most labs do have webpages that contain their publications. Google scholar is pretty good at finding them too. This covers ~99.9% of cases, and they're all instant once you know how to use them. My point in the original comment was if that these fail, you should shoot the authors an email instead of paying. A few publishers do hold a fairly hard line against publicly posting an article, but even they typically allow you to provide a copy "on request". Also, in case it's not clear, academic publishing is only partly about actually disseminating information. A huge part is getting the "imprimatur" of peer review generally, and from certain journals specifically. A brilliant self-hosted document might impress people within a very specific subfield that are motivated and able to evaluate the work. However, its value will be unclear to others who fund/hire/evaluate you. The Dean, for example, likely has neither has the background nor the time to decide if your technical blog merits tenure. He/she does know that Nature is very selective, publishing there means that your work is held in high regard by your field, and therefore you deserve tenure. This holds true for funders too. It's not a great system in many regards, but it is how things currently work.