- more and more research institutes have a public repository of articles.
- plenty of authors put their own (preprints of) papers on their personal website, a preprint server like arxiv, or researchgate.
- lastly, you can mail an author and request a preprint. Not all will reply, but plenty will be happy that someone noticed their work and send a copy.
For some reason, they're also slow to respond to emails.
For any papers published before 1965, even with maximal renewals available at the time of publication, the works would now be in the public domain under copyright law at the time of publication. Otherwise, many works published prior to 1976 which were either only renewed once or never registered would similarly have lapsed to public domain, or always been so.
That's a huge effectively-lost trove of knowledge.
For a single paper, googling for the pdf, failing that, searching for author page / institutional repo doesn't take much effort. A few minutes, I'd guess.
That still becomes prohibitive for >10 papers sought. Then you need institutional access, something like Sci-Hub, or a lot of patience.
Mailing works as a last resort option.
Honestly, part of the reason i jump through such hoops is to justify not including a paper. I have institutional access and all these options. If none of that suffices, I feel justified in considering the alleged article as not part of the scientific record.
That's a fraction of the citations on most interesting papers. I'm not a practicing scientist (now or ever) so my access patterns are different but every time I explore a topic in academic journals that process becomes prohibitively expensive almost immediately. Even when I had institutional access it was a pain in the ass - without SciHub every doi is a game of subscription roulette.
Asking by email.. too many publications at the moment.