01Start with the one nobody uses.
Interlibrary loan. Almost every institution runs it, most online students have never heard of it, and it is the single most useful service a library offers. You submit the citation, somebody obtains a copy from another library, and it arrives by email.
It is usually free to the student, frequently faster than expected, and it works for chapters and whole books as well as articles. There is often a limit per term, which is worth knowing so it can be spent on the sources that genuinely matter rather than on everything you might read.
The reason to try it first is that it goes to work while you carry on. Submit the request, keep writing around the gap, and deal with it when the file lands. Waiting to see whether a paper arrives before continuing is what turns a small obstacle into a lost week.
02The other five, in order.
Each takes a few minutes, and one of them nearly always works.
- Check whether your library holds it under a different route. Databases overlap unevenly, and a paper unavailable through one may sit in another your institution also subscribes to. The library's own discovery search covers all of them at once.
- Look for an author-deposited version. Many funders and institutions require a copy to be placed in an open repository, which is legal, free, and usually the accepted manuscript rather than the typeset one.
- Check the journal's own age policy. A good many journals open their archives after a period, so a paper from some years back is often free at source without anybody advertising it.
- Write to the author. This works far more often than students believe and is entirely normal academic conduct.
- And a public library card, which in many places carries database access an institution's own subscription does not cover.
03Writing to an author.
This is the route students find strangest and it is genuinely ordinary. Academics are usually pleased that somebody wants to read their work, they are permitted to share their own manuscript on request, and the whole exchange takes them under a minute.
Keep it to three lines. Who you are and what you are studying, which paper you are after by full title, and a direct request for a copy. No flattery, no long explanation of your project, and no apology for asking. The email address is nearly always on the paper's own landing page or on a departmental profile.
Expect a reasonable proportion to go unanswered, because inboxes are inboxes, and send a couple rather than agonising over one. The tone to aim for is close to what the issue on writing to a professor describes: short enough to answer between two other things.
04When to stop chasing one paper.
Sooner than most people do. There is a particular kind of stuck where somebody spends three days trying to obtain one article because a reference list implied it was important, and the paper turns out to be tangential when it finally arrives.
Before spending more than about twenty minutes, ask what the paper is actually doing in your argument. If something else could support the same claim, use that. If the paper is genuinely load-bearing, request it properly through interlibrary loan and keep writing around the gap while it comes.
There is also a decent chance the thing you need is downstream rather than upstream: a more recent paper citing it will often summarise what you were after and be freely available, which is the technique described at citation chaining.
05The route not to take.
Sites offering paywalled research free at scale exist and are widely used, and this house is not going to pretend otherwise or lecture anybody about it. What is worth saying plainly is the practical position rather than the moral one.
The legal status is contested and varies by jurisdiction. Some institutions block them and log the attempt. And the copies are of unknown provenance, which matters for the reason set out in the retractions issue: a file obtained outside the publisher's record carries no indication of whether the paper has since been corrected or withdrawn.
Given that interlibrary loan is free, legal, and gets you the same article with its status intact, the calculation is not close for anybody who has been shown the alternative. Most people simply never were.
Questions at this desk.
How long does interlibrary loan take?
Frequently days rather than weeks for an article, and longer for a book, though it varies by institution and by how obscure the item is. The practical implication is to request early and keep working, rather than treating the request as a reason to pause.
Is it acceptable to email an author?
Entirely, and it is ordinary academic conduct rather than an imposition. Authors are permitted to share their own manuscripts on request and most are glad to. Keep it to three lines, expect some silence, and send a couple rather than investing hope in one.
What if my library holds almost nothing?
Then interlibrary loan matters more, not less, and it exists precisely for that situation. A public library card is also worth having alongside your institutional access, since the two cover different things and the combination is often broader than either.
Should I ever pay for an article myself?
Rarely, and not before the free routes have been tried. If a single paper is genuinely load-bearing for a thesis and every route has failed, that is a defensible purchase. Paying for coursework reading because nobody mentioned interlibrary loan is not.
Notes in the margin, from the people at the desks.
READ THE LETTER