{"course":{"slug":"ai-for-students","title":"AI for Students","url":"https://addaly.com/learn/ai-for-students","level":"beginner","track":"work"},"lesson":{"slug":"finding-the-paper-free","title":"Finding the paper, for nothing","url":"https://addaly.com/learn/ai-for-students/finding-the-paper-free","position":40,"of":73,"minutes":9,"answer":"Library proxy, Unpaywall, Scholar's all-versions, preprint servers, PMC and emailing the author will reach almost any paper free; and the page number you record while reading is what turns a citation into evidence.","keyIdea":"Library proxy, Unpaywall, Scholar's all-versions, preprint servers, PMC and emailing the author will reach almost any paper free; and the page number you record while reading is what turns a citation into evidence."},"body":{"format":"markdown","text":"## Paywalls are navigable\n\nA great deal of research is behind subscription. Very little of it is unreachable if you know the routes, and all of the routes below are legal and free.\n\n**Your library first.** Institutional access covers most of what you need, and the route is usually a proxy login on the library site rather than the publisher's page directly. If you hit a paywall, go through the library catalogue rather than Google, and it often opens. Librarians will also request anything the library does not hold, through inter-library loan, usually free for students and usually within days. This service is heavily underused.\n\n**Unpaywall.** A free browser extension that checks, for any article page, whether a legal open-access copy exists: in a repository, a preprint server, or the author's own site. It finds one surprisingly often. **Open Access Button** does the same job.\n\n**Google Scholar.** Search the title in quotation marks. Scholar shows links to free versions on the right-hand side, and *All versions* under a result often surfaces a PDF in a university repository.\n\n**Preprint servers.** arXiv for physics, maths, computer science and increasingly other fields; bioRxiv and medRxiv for life sciences and medicine; SSRN for social science and law; PsyArXiv for psychology. The preprint is frequently the same content as the published paper, minus copy-editing.\n\n**PubMed Central**, free full text for a large share of biomedical literature, and required for research funded by many public bodies. **Europe PMC** covers similar ground.\n\n**Institutional repositories.** Most universities require deposit of their staff's papers. Search the author's institution repository directly.\n\n**Email the author.** This works, it is normal, and researchers are usually pleased. Two sentences: who you are, which paper, that you cannot access it, and would they mind sending a copy. Their email is on the paper.\n\n## Books\n\n**Internet Archive** lends scanned books free, one borrower at a time, and has an enormous collection. **Project Gutenberg** and **Standard Ebooks** for out-of-copyright works. **HathiTrust** for full-text search across scanned books, which lets you find *which page* something is on even when you cannot read the whole book; then use your library for that page. **Google Books** allows preview of many pages and, crucially, full-text search, which is how you verify a quotation.\n\nAnd the physical library. The book is on a shelf, it is free, and nobody else is looking at it.\n\n## Verifying in the right order\n\nA cheap sequence, roughly three minutes per source:\n\n1. Search the exact title in quotation marks. Nothing? Suspect fabrication.\n2. Resolve the DOI. Fails? Fabrication or a wrong DOI.\n3. Open the abstract. Does it look like the paper your claim needs?\n4. Get the full text through one of the routes above.\n5. Find the specific sentence, figure or table your claim rests on. Note the page number.\n\nStep five is where the citation becomes yours rather than something you inherited. It is also what makes the writing better, because a page number forces you to say what the source actually shows rather than what it is generally about.\n\n## Search that finds things\n\nDatabase search is a skill and it is not the same as asking a question. What works:\n\n- **Quotation marks** for exact phrases.\n- **Boolean operators** (`AND`, `OR`, `NOT`) supported in most academic databases.\n- **Field limits.** `author:`, `title:`, date ranges. Scholar and PubMed both support these.\n- **Following citations.** Find one good paper, then look at its reference list for the older work and Scholar's *Cited by* for the newer. This is the most efficient search method there is and it uses somebody else's judgement about relevance.\n- **Review articles first.** A recent review gives you the map of a field and a curated reference list. Search for `[topic] systematic review` or `[topic] narrative review`.\n\nA model is useful for step zero here: turning your question into the vocabulary a database uses. Fields have terms of art, and searching for the plain-English version returns nothing. Ask what the technical terms are, then search for those.\n\n## Managing what you find\n\n**Zotero** is free, open source, and better than the paid alternatives. Browser plugin saves a reference and the PDF in one click, generates bibliographies in any style, and syncs. Install it in your first week and never assemble a reference list by hand again.\n\nOne caution that comes up constantly: Zotero's automatic metadata is usually right and occasionally wrong, especially for chapters and older material. Check what it captured. A reference list where every entry is subtly malformed is a different kind of avoidable mark loss."},"check":{"question":"Why is following a good paper's reference list and its \"cited by\" list often more efficient than searching a database directly?","options":[{"text":"Because citation chains avoid the paywall problem, since cited papers are more likely to be open access.","correct":false,"tempting":"Invents a correlation; whether a cited work is open depends on its publisher, not on having been cited."},{"text":"Because databases index only a fraction of published work, whereas reference lists capture everything in a field.","correct":false,"tempting":"Reverses the coverage relationship; major databases index far more than any single paper cites."},{"text":"Because reference lists are curated by the paper's authors, so you inherit an expert's judgement about what is relevant.","correct":true},{"text":"Because it removes the need to know a field's technical vocabulary, which is the main obstacle to effective database search.","correct":false,"tempting":"Names a real obstacle the lesson raises but attaches it to the wrong solution; reading reference lists is in fact how you acquire that vocabulary."}]},"prev":{"slug":"how-a-citation-gets-invented","title":"How a citation gets invented","url":"https://addaly.com/learn/ai-for-students/how-a-citation-gets-invented"},"next":{"slug":"reading-a-paper-fast","title":"Reading a paper without reading all of it","url":"https://addaly.com/learn/ai-for-students/reading-a-paper-fast"},"authorship":"Founder and series editor: Dr Mustafa Mun. Written by AI models to a written standard. Exam keys checked blind by a second model: 3,665 of 3,666 agreed. Academic review invited.","review":{"status":"Draft, open for academic review"},"cite":"Addaly (2026) 'Finding the paper, for nothing', in AI for Students. Edited by M. Mun. Addaly. Available at: https://addaly.com/learn/ai-for-students/finding-the-paper-free (Accessed: 13 October 2026).","licence":{"name":"CC BY-NC-SA 4.0","url":"https://creativecommons.org/licenses/by-nc-sa/4.0/"},"reader":{"schema":1,"title":"Finding the paper, for nothing","courseTitle":"AI for Students","module":{"index":5,"slug":"research-and-evidence","title":"Research, sources and evidence"},"answer":null,"keyIdea":"Library proxy, Unpaywall, Scholar’s all-versions, preprint servers, PMC and emailing the author will reach almost any paper free; and the page number you record while reading is what turns a citation into evidence.","hero":{"kind":"plate","image":null,"alt":null,"caption":null},"blocks":[{"t":"h2","id":"paywalls-are-navigable","runs":[{"text":"Paywalls are navigable"}]},{"t":"p","runs":[{"text":"A great deal of research is behind subscription. Very little of it is unreachable if you know the routes, and all of the routes below are legal and free."}]},{"t":"p","runs":[{"text":"Your library first.","b":true},{"text":" Institutional access covers most of what you need, and the route is usually a proxy login on the library site rather than the publisher’s page directly. If you hit a paywall, go through the library catalogue rather than Google, and it often opens. Librarians will also request anything the library does not hold, through inter-library loan, usually free for students and usually within days. This service is heavily underused."}]},{"t":"p","runs":[{"text":"Unpaywall.","b":true},{"text":" A free browser extension that checks, for any article page, whether a legal open-access copy exists: in a repository, a preprint server, or the author’s own site. It finds one surprisingly often. "},{"text":"Open Access Button","b":true},{"text":" does the same job."}]},{"t":"p","runs":[{"text":"Google Scholar.","b":true},{"text":" Search the title in quotation marks. Scholar shows links to free versions on the right-hand side, and "},{"text":"All versions","i":true},{"text":" under a result often surfaces a PDF in a university repository."}]},{"t":"p","runs":[{"text":"Preprint servers.","b":true},{"text":" arXiv for physics, maths, computer science and increasingly other fields; bioRxiv and medRxiv for life sciences and medicine; SSRN for social science and law; PsyArXiv for psychology. The preprint is frequently the same content as the published paper, minus copy-editing."}]},{"t":"p","runs":[{"text":"PubMed Central","b":true},{"text":", free full text for a large share of biomedical literature, and required for research funded by many public bodies. "},{"text":"Europe PMC","b":true},{"text":" covers similar ground."}]},{"t":"p","runs":[{"text":"Institutional repositories.","b":true},{"text":" Most universities require deposit of their staff’s papers. Search the author’s institution repository directly."}]},{"t":"p","runs":[{"text":"Email the author.","b":true},{"text":" This works, it is normal, and researchers are usually pleased. Two sentences: who you are, which paper, that you cannot access it, and would they mind sending a copy. Their email is on the paper."}]},{"t":"h2","id":"books","runs":[{"text":"Books"}]},{"t":"p","runs":[{"text":"Internet Archive","b":true},{"text":" lends scanned books free, one borrower at a time, and has an enormous collection. "},{"text":"Project Gutenberg","b":true},{"text":" and "},{"text":"Standard Ebooks","b":true},{"text":" for out-of-copyright works. "},{"text":"HathiTrust","b":true},{"text":" for full-text search across scanned books, which lets you find "},{"text":"which page","i":true},{"text":" something is on even when you cannot read the whole book; then use your library for that page. "},{"text":"Google Books","b":true},{"text":" allows preview of many pages and, crucially, full-text search, which is how you verify a quotation."}]},{"t":"p","runs":[{"text":"And the physical library. The book is on a shelf, it is free, and nobody else is looking at it."}]},{"t":"h2","id":"verifying-in-the-right-order","runs":[{"text":"Verifying in the right order"}]},{"t":"p","runs":[{"text":"A cheap sequence, roughly three minutes per source:"}]},{"t":"ol","start":1,"items":[[{"text":"Search the exact title in quotation marks. Nothing? Suspect fabrication."}],[{"text":"Resolve the DOI. Fails? Fabrication or a wrong DOI."}],[{"text":"Open the abstract. Does it look like the paper your claim needs?"}],[{"text":"Get the full text through one of the routes above."}],[{"text":"Find the specific sentence, figure or table your claim rests on. Note the page number."}]]},{"t":"p","runs":[{"text":"Step five is where the citation becomes yours rather than something you inherited. It is also what makes the writing better, because a page number forces you to say what the source actually shows rather than what it is generally about."}]},{"t":"h2","id":"search-that-finds-things","runs":[{"text":"Search that finds things"}]},{"t":"p","runs":[{"text":"Database search is a skill and it is not the same as asking a question. What works:"}]},{"t":"ul","items":[[{"text":"Quotation marks","b":true},{"text":" for exact phrases."}],[{"text":"Boolean operators","b":true},{"text":" ("},{"text":"AND","code":true},{"text":", "},{"text":"OR","code":true},{"text":", "},{"text":"NOT","code":true},{"text":") supported in most academic databases."}],[{"text":"Field limits.","b":true},{"text":" "},{"text":"author:","code":true},{"text":", "},{"text":"title:","code":true},{"text":", date ranges. Scholar and PubMed both support these."}],[{"text":"Following citations.","b":true},{"text":" Find one good paper, then look at its reference list for the older work and Scholar’s "},{"text":"Cited by","i":true},{"text":" for the newer. This is the most efficient search method there is and it uses somebody else’s judgement about relevance."}],[{"text":"Review articles first.","b":true},{"text":" A recent review gives you the map of a field and a curated reference list. Search for "},{"text":"[topic] systematic review","code":true},{"text":" or "},{"text":"[topic] narrative review","code":true},{"text":"."}]]},{"t":"p","runs":[{"text":"A model is useful for step zero here: turning your question into the vocabulary a database uses. Fields have terms of art, and searching for the plain-English version returns nothing. Ask what the technical terms are, then search for those."}]},{"t":"h2","id":"managing-what-you-find","runs":[{"text":"Managing what you find"}]},{"t":"p","runs":[{"text":"Zotero","b":true},{"text":" is free, open source, and better than the paid alternatives. Browser plugin saves a reference and the PDF in one click, generates bibliographies in any style, and syncs. Install it in your first week and never assemble a reference list by hand again."}]},{"t":"p","runs":[{"text":"One caution that comes up constantly: Zotero’s automatic metadata is usually right and occasionally wrong, especially for chapters and older material. Check what it captured. A reference list where every entry is subtly malformed is a different kind of avoidable mark loss."}]}],"check":{"question":[{"text":"Why is following a good paper’s reference list and its “cited by” list often more efficient than searching a database directly?"}],"options":[{"text":[{"text":"Because citation chains avoid the paywall problem, since cited papers are more likely to be open access."}],"correct":false,"tempting":[{"text":"Invents a correlation; whether a cited work is open depends on its publisher, not on having been cited."}]},{"text":[{"text":"Because databases index only a fraction of published work, whereas reference lists capture everything in a field."}],"correct":false,"tempting":[{"text":"Reverses the coverage relationship; major databases index far more than any single paper cites."}]},{"text":[{"text":"Because reference lists are curated by the paper’s authors, so you inherit an expert’s judgement about what is relevant."}],"correct":true,"tempting":null},{"text":[{"text":"Because it removes the need to know a field’s technical vocabulary, which is the main obstacle to effective database search."}],"correct":false,"tempting":[{"text":"Names a real obstacle the lesson raises but attaches it to the wrong solution; reading reference lists is in fact how you acquire that vocabulary."}]}],"order":[1,0,2,3],"why":null},"prev":{"slug":"how-a-citation-gets-invented","title":"How a citation gets invented","minutes":8},"next":{"slug":"reading-a-paper-fast","title":"Reading a paper without reading all of it","minutes":8}}}