Reader Guidelines ================= Tillyard. Version 1.0. In effect from October 20, 2026. Fingerprint of this text (SHA-256): d3d32e6372a4ef98133d084dbf8e9f833b10a38dc47fcbbda57abb8ace6b916e This version: https://tillyard.org/policies/reader-guidelines/v/1.0 These guidelines apply to members who join the reading pool of Tillyard Research. You agree to them when you join the pool, and you confirm the two most important points again each time you take a paper. They are public so that authors, and anyone reading a published paper, can see what readers were asked to do. - Reading is your choice, every time. You can join, pause or leave the pool in one click, and say no to any paper. Declining never counts against you. - Nothing is given for reading: no pay, no perks, no standing, and no faster handling of your own paper. Nothing is taken away for not reading. - An unpublished paper is confidential. Do not share it, copy it, keep it, or put any part of it into an outside tool, including an AI tool. - If you know or can guess who wrote a paper, or you have something to gain from its result, step aside. - Score six rubric items from 0 to 2 and give a reason for each. Open two of the paper's sources yourself and write down what you found. - Write for the author: specific, fair and about the paper. Do not guess who they are, and do not make accusations. - What you read is not yours to use. The author's ideas, data, sources and wording stay theirs. - You give an opinion. The editor decides. Readers are not staff. If this summary and the full text below ever differ, the full text governs. 1. What reading is ================== 1.1 What a reader does ---------------------- A reader is a member who has chosen to read a paper that another member has sent to Tillyard Research, to score it against the published rubric, to check two of its sources, and to write comments the author can use. The reader does not know who wrote the paper, and the author never learns who read it. The readings go to the editor, who decides whether the paper is published. 1.2 It is voluntary ------------------- Nobody is required to read. You join the reading pool because you want to. Once you are in it, we may offer you a paper now and then; you may take it or decline it. You can pause offers or leave the pool at any time, each with one click. No number of readings is needed for anything on the site. 1.3 It is unrewarded -------------------- You are not paid for reading. Reading earns no perk, stamp, title or standing. It does not move your own paper up the line, and it is not a condition of having your own paper read. Declining, pausing, leaving, or letting a date go by costs you nothing. We do not ask members to do required work for us. So reading stays a choice, with nothing hanging on it. 1.4 What you may keep --------------------- Two small things follow from a reading, and neither is a reward. First, your own Record can hold a private count of the papers you have read (for example "Read 6 student papers for Tillyard Research"). It states a fact, it is yours to show or to keep private, and it is off until you turn it on. Second, an author may press "This helped" on a reading; you then receive a thanks with no names attached. It carries nothing with it. 1.5 Readers are not staff ------------------------- A reader gives an opinion. Readers do not decide what is published, do not moderate, and do not act or speak for Tillyard. Reading does not make you an employee, intern, contractor, agent or volunteer of Tillyard. Nobody supervises your reading, sets your hours or gives you a target. The limits on how many papers you can hold at once exist to protect authors' work; they are ceilings and never goals. 1.6 Where else these guidelines apply ------------------------------------- Some tracks in Pursuits offer members of the same pool a portfolio piece to read, and ask those who take it to type what they observed. Sections 3 to 8 and 10 of these guidelines apply to that reading too, alongside the track's own published rules. Two other arrangements are different and have their own rules: a contest round whose published rules say that entrants score other entries, and a swap that two members agree between themselves (such as the résumé exchange in The Workbench). 1.7 How this fits with our other documents ------------------------------------------ The Editorial Policy of Tillyard Research describes how papers are handled and decided. The Author Agreement and Licence sets out the author's rights. The AI-Use Policy covers AI tools. The Community Guidelines and the Terms of Use cover conduct on the site in general, and the Privacy Notice covers what we keep. On what a reader must and must not do with a paper, these guidelines govern because they are the more specific document. Section 37.5 of the Terms of Use sets the order among all our documents, including the sections of those terms that prevail over every other document. Words with a set meaning across our documents, such as member, staff, the editor, reader, hold and case, are defined in Definitions. 1.8 Numbers that can change --------------------------- Where these guidelines give a number and say "currently", that number is a setting we can change. The number shown on the page when you act is the one that applies. 2. Joining the pool and how papers reach you ============================================ 2.1 Who can join ---------------- You can join the pool if you are a member in good standing whose account is old enough (currently at least 14 days). Before joining you read the reader's guide and do one practice reading on a specimen paper. Tillyard wrote the specimen; no member submitted it. Afterwards you see the editor's scoring and reasons beside your own. There is no mark, the result is private to you, and it never appears on The Record. 2.2 Choosing fields ------------------- You tell us which fields you are comfortable reading. For the first few days a paper is offered only to readers who listed its field (currently three days); after that it may be offered to any reader in the pool. 2.3 You never choose the paper ------------------------------ Papers are drawn at random. You cannot pick one, ask for one, or arrange to receive one. The draw leaves out anyone who is an author of the paper, anyone at an author's college, anyone connected with or blocked by an author on the site, anyone who has been offered that paper before, and anyone who in the last year has read a paper by one of its authors or read alongside one of its other readers. 2.4 The offer ------------- When a paper is drawn for you, you get a notice and a card on your home page. The offer tells you the kind of paper, its field, its length and an estimate of the reading time. It does not tell you the title or anything about the author. You have a set time to answer (currently 48 hours). Press "Read this paper" or "Not this one". If you do nothing, the offer lapses and the paper goes to someone else. 2.5 Taking a paper ------------------ Before the text is shown, you confirm two things: "I do not know who wrote this and have nothing to gain from its result." and "I will keep it to myself." Section 3 explains the first and section 4 the second. You then have a return-by date (currently seven days). You can add time once yourself (currently three more days). 2.6 While you hold it --------------------- The paper's text is available to you only between taking it and returning it. You can save a partly finished reading and come back to it. There is no timer on the form. 2.7 Ceilings ------------ To protect authors' work, a reader can hold only a small number of papers at once (currently one) and take only a few in a week (currently three). These are limits on how much you can take on. They are never a target. 3. Conflicts of interest ======================== 3.1 What a conflict is ---------------------- You have a conflict if you know or can guess who wrote the paper, or if you have something to gain or lose from what happens to it. A reader with a conflict cannot give the author a fair reading, and the wording on a published paper ("Read by three students at other colleges") would be less true than it should be. 3.2 What the draw already rules out ----------------------------------- The draw keeps a paper away from its authors, from members at the authors' colleges, and from members connected with the authors on the site (see 2.3). It cannot know what you know. 3.3 What only you can know -------------------------- Step aside if any of these is true: 1. you recognize the paper, the project, the data or the writing, or you have seen a draft of it before; 2. you think you know who wrote it, from the topic, the place it describes, or anything else; 3. the author may be a friend, a relative, a partner or former partner, a teammate, a classmate from an earlier college, or someone you have fallen out with; 4. you are working on the same question yourself, or have a paper of your own on a closely related topic in the line or in preparation; 5. you have a stake in the subject: you work for, belong to, or are in a dispute with an organization the paper is about, or you would gain or lose if its conclusions were accepted; 6. you hold views on the subject so strong that you doubt you can score the paper on the rubric alone. 3.4 How to step aside --------------------- If you have not taken the paper, press "Not this one". If you have taken it, press "I think I know whose this is". The paper goes to another reader. You are not asked who you think wrote it, and you should not say. Do not try to confirm a guess by searching for the paper's subject, its phrases or its author. 3.5 If you realize later ------------------------ If you realize part-way through, step aside in the same way. If you realize after sending your reading, tell the editor at support@tillyard.org. The editor can set a reading aside. 3.6 When in doubt ----------------- Step aside. Nobody asks why, and it is never counted. 3.7 After the reading --------------------- Do not try to find out who wrote a paper you read, and do not contact them about it. If you later meet the author or connect with them on the site, do not tell them you read their paper. Do not tell anyone else which papers you read, before or after publication. You may say in general that you have read for Tillyard Research, and how many papers. 4. Confidentiality of unpublished drafts ======================================== 4.1 Everything about the paper is confidential ---------------------------------------------- An unpublished paper belongs to its author, and you are seeing it only because you agreed to keep it private. That covers its text, title, topic, question, method, data, findings and sources. It covers the fact that the paper exists and that you read it. It covers your own scores and comments, and anything you learn about what was decided. 4.2 What you must not do with it -------------------------------- Do not: 1. show, send, read out or describe the paper to anyone, including friends, classmates, tutors and professors; 2. post anything about it, anywhere, including in a table chat, a note or a thread on this site; 3. copy it, download it, photograph it, take a screenshot of it or print it; 4. keep any part of it after you return your reading; 5. read it where someone else can see your screen. 4.3 No outside tools -------------------- Do not paste, upload or type any part of the paper into another service. That includes AI assistants and chatbots, translation services, grammar and paraphrasing tools, plagiarism checkers, AI detectors, shared documents and note-taking apps that store text online. Sending an author's unpublished words to another company's servers is sharing them. To check a source, open the source; do not paste the author's sentence into a search box. 4.4 Assistive technology ------------------------ The technology you ordinarily use to read and write, such as a screen reader, magnification or dictation, is fine. 4.5 Your own notes ------------------ Write your reading in the form on the site. If you made private notes while reading, delete or destroy them once you have sent your reading. 4.6 How long this lasts ----------------------- The duty lasts until the paper is published. If the paper is never published, it does not end. After publication the published text is public, and you may read and cite it like anyone else. The draft you saw, the ways it differs from the published version, your own reading and the fact that you were a reader stay confidential. 4.7 When you may speak ---------------------- You may always tell us about a paper: in the note for the editor, through the report control, or at support@tillyard.org. Nothing in these guidelines stops you from reporting a crime to the authorities, from getting emergency help for someone who seems to be in danger, or from answering a court order or other legal demand. If you receive a legal demand about a paper you read, please tell us first if the law allows you to. 4.8 If something goes wrong --------------------------- If you shared something by mistake, or you think someone else has seen the paper through you, tell us at once. We would much rather hear it from you. 5. No use of another member's unpublished work ============================================== 5.1 Do not use what you learn from an unpublished paper ------------------------------------------------------- Do not use anything you learn from an unpublished paper for your own purposes. Reading it does not make any of it yours. 5.2 What that covers -------------------- It covers the author's research question, their idea for a study, their method, their data and results, their argument and conclusions, the way they brought their sources together, the structure of the paper and its wording. It applies to your own papers, your coursework, your projects, your applications and anything else you make. 5.3 Once the paper is published ------------------------------- You may then use it as you would any published source: read the published version and cite it. 5.4 If the paper is never published ----------------------------------- The rule continues. You may not use what you read. 5.5 Work you had already begun ------------------------------ If a paper turns out to cover something you were already working on by yourself, step aside (section 3) and tell the editor. Keep your own dated notes and drafts; they show what was yours beforehand. 5.6 Published sources you first met in a draft ---------------------------------------------- A journal article or a book that the author cited is a public source, and nothing stops you from reading it. What you must not take is the author's use of it: their selection of sources, their reading of them and the argument they built. 5.7 No advantage of any kind ---------------------------- Do not use what you learn as a reader to get ahead of the author or anyone else, for example in a contest round, an application or a class. 6. How to read and score ======================== 6.1 Read it through first ------------------------- Read the whole paper once before you score anything. Each kind of paper (a research report, a literature review, a case study, a data note, a policy memo, a replication or fact-check, a profile) has its own outline, and two rubric items that matter most for that kind. The form shows a prompt for those two. 6.2 The six items ----------------- You score six items. Each takes 0 ("Not yet"), 1 ("Partly") or 2 ("Clearly"). The rubric published on the About page of Tillyard Research, with its description of what each score looks like, is the governing text. The table below is our guidance on applying it. The six items. Columns: item, the question to ask, 0, not yet, 1, partly, 2, clearly. Item | The question to ask | 0, not yet | 1, partly | 2, clearly ---------------------------------------------------------------- Clear question | Does the paper say early what it is trying to find out or argue, and stay with it? | No question can be found, or the paper wanders between several | A question is there, but it is vague, too broad, or the paper drifts from it | A specific question is stated early and the paper stays on it Method described | Could you follow what the author did from their description? | No account of how the author went about it | Some steps are given; important choices are missing (which sources, which data, how they were picked) | The steps and choices are clear enough that another student could repeat them Evidence traceable | Can each main claim be followed to something you could check? | The main claims rest on assertion | Some claims can be traced and others cannot | Each main claim points to a source, a table or data that you could find Sources real and relevant | Do the sources exist, say what the paper says they say, and suit the claims they support? | Sources cannot be found, or do not say what is claimed | The sources exist, but several are weak, off the point or second-hand | The sources exist, support the claims and fit the question The writer's own analysis | Does the author do something with the material: compare, weigh, explain, conclude? | Summary or a string of quotations only | Some analysis, but mostly a restatement of sources | The author's own reasoning is visible and carries the paper Limitations stated | Does the paper say what it cannot show? | Nothing | A passing or generic mention ("more research is needed") | Specific limits are named, with what they mean for the conclusion 6.3 Score each item by itself ----------------------------- Give each item the score it deserves, whatever the others get. Do not work backward from an overall impression, and do not try to guess what other readers will say. There is no total that a paper has to reach, and your total is not a mark. Scores are never shown publicly. 6.4 What the rubric does not ask -------------------------------- Do not score the paper on whether you agree with it, on whether the topic interests you, on how polished or fluent the English is, on its length within the range for its kind, or on anything you imagine about the author. A clearly written paper with a clear question and checkable evidence scores well. A fluent paper without them does not. 6.5 If it is outside your field ------------------------------- You are a student reader, and nobody expects a specialist's view. Score what you can judge and say what you cannot: "I cannot assess the statistics. The method section does tell me how the sample was chosen." That is a useful sentence for the author and for the editor. 6.6 Give a reason for every score --------------------------------- Each score needs a sentence or two saying why (the form sets a minimum length). Point to the place in the paper: "Section 3 says prices rose 40% but gives no source." Write a different reason for each item; the form will not accept the same reason twice. 6.7 One thing done well, and the first thing to fix --------------------------------------------------- The form asks for both. Make the strength real and particular. Make the fix the single change that would help the paper most, and say how the author might go about it. 6.8 The source check -------------------- We draw two of the paper's sources. Open each one yourself and choose what you found: "It says what the paper says", "It says something different", or "I could not open it". Then quote the line you found, or say what you found instead. Typing what you saw is the point of the check; a tick alone would tell nobody anything. 1. "I could not open it" is an honest and acceptable answer. Some sources are behind paywalls or are books you do not have. 2. Use your own college library's access if you have it. Do not use pirate sites, and do not ask anyone to send you the source in a way that would reveal what you are reading. 3. If a source seems not to exist, say exactly what you tried: "The DOI leads to a different article, titled ...". 6.9 A note for the editor only ------------------------------ This box is optional. Use it for things the author should not read from you directly: a conflict you noticed late, a worry that part of the paper is copied, a source that looks invented, personal details about someone that should not be there, or a safety concern. Write what you observed, and leave conclusions to the editor. "Paragraphs 4 to 6 match the opening of the Wikipedia article on this topic almost word for word" is an observation. "This is plagiarism" is a verdict. The author is not shown this note. It is kept with the paper's records, and like any record it could have to be produced if the law required it. 6.10 Sending ------------ When you press "Send reading", your scores are final and cannot be changed. If you realize you made a mistake, write to support@tillyard.org; the editor can take that into account or set the reading aside. 7. Writing comments the author can use ====================================== 7.1 Who you are writing to -------------------------- Another student, who chose their own topic, worked on it for weeks, and will read every word you write. They cannot reply to you. Write what you would want to receive. 7.2 What useful comments have in common --------------------------------------- They are specific, they point to a place in the paper, they explain the problem and they suggest a way forward. What useful comments have in common. Columns: less useful, more useful. Less useful | More useful ------------------------- "The method is weak." | "The method section says you surveyed students, but not how many, how you reached them, or when. Adding those three facts would let a reader judge the results." "Needs more sources." | "The claim in section 2 that food banks lose a fifth of donated produce rests on one news article. Is there a report or a study behind that article you could cite directly?" "Good paper." | "The comparison of the two counties in section 4 is the strongest part: it is the one place where your own data answers your question." "I don't buy the conclusion." | "The conclusion says the policy caused the drop, but section 3 shows the drop began a year earlier. Could you address that, or soften the claim?" "Confusing." | "I lost the thread in section 3, where the paper moves from interviews to national statistics without saying why. A sentence linking the two would help." 7.3 Tone -------- Be direct and be kind. Write about the paper ("the paper claims", "section 2 does not say") and leave the person out of it. No sarcasm, no jokes at the author's expense, and no remarks about how much work they did or did not do. 7.4 Nothing about who the author is ----------------------------------- Do not guess at or comment on the author's identity, college, year, background, first language, gender, race, religion, politics, disability or anything else about them. Do not treat unpolished English as a sign of anything. If a sentence is unclear, say which one and what you could not follow. 7.5 No accusations ------------------ Do not tell the author that their paper is copied, invented or written by AI. You cannot know that from reading it, and a wrong accusation does real harm. Describe what you can see: "Section 3 makes four factual claims with no source", or "The source for the 2019 figure says 2009". If you have a concern that goes beyond what you can see, put it in the note for the editor, as an observation. 7.6 No verdicts and no marks ---------------------------- Do not write "accept", "reject", "publish", or a letter or number mark of your own. Readers score the six items. The editor decides. 7.7 Nothing that identifies you ------------------------------- Leave out anything that could tell the author who you are: your college, your major, your own paper or project, your classes, your name, your contact details or a link. Do not ask the author to cite a particular piece of your own work. 7.8 Do not rewrite the paper ---------------------------- Point to what needs work and why. The paper is the author's to write. 7.9 Your words are screened --------------------------- Your comments pass through the same automated text-safety screening as other member text before the author can see them. If something is held, a person looks at it. 8. What readers must not do =========================== For ease of reference, here is the whole list in one place. A reader must not: 1. take a paper when they know or can guess who wrote it, or have something to gain from its result (section 3); 2. try to find out who wrote a paper, or contact its author, its co-authors, the people it names or the author's college; 3. tell anyone which paper they read, or tell an author that they were the reader (3.7); 4. share, copy, keep or describe an unpublished paper (section 4); 5. put any part of an unpublished paper into an outside service, including an AI tool, a plagiarism checker or an AI detector (4.3); 6. use an AI tool to read, score or judge a paper, or to write the reading. Your reading must be your own judgment in your own words. A spelling or grammar check on your own comments is fine so long as none of the paper's text goes into it; 7. use anything from an unpublished paper in their own work (section 5); 8. send a reading they did not do: scoring without reading the paper, repeating one reason across items, reusing the text of an earlier reading, or filling in a source check without opening the source; 9. arrange, or try to arrange, who reads whose paper: by trading readings with a friend, by using a second account, or by asking others to join the pool in the hope of drawing a particular paper; 10. accept or offer anything in return for a reading or a score: money, a favor, a tip, or a kinder reading in return; 11. use a reading to insult, belittle, threaten or harass, or to settle a score (section 7); 12. make accusations about the author in comments the author will read (7.5); 13. describe their reading as more than it is. On a résumé or profile, use the wording on your Record ("Read 6 student papers for Tillyard Research"). Do not call yourself a peer reviewer for a journal, a referee, an editor or a member of an editorial board, do not use any title that suggests you worked for Tillyard, and do not list reading under work experience; 14. speak or act for Tillyard. 9. Declining, pausing and leaving ================================= 9.1 Declining an offer ---------------------- Press "Not this one". You do not give a reason. The paper goes straight to another reader. 9.2 Letting an offer lapse -------------------------- If you do not answer in time, the offer ends and the paper goes to someone else. That is all that happens. 9.3 After you have taken a paper -------------------------------- If you find you cannot finish, you do not need to do anything. When the return-by date goes by, the paper quietly goes to another reader, and you see: "This paper has gone to another reader. Nothing counts against you." If you step aside because you may know the author, use "I think I know whose this is" (3.4). A reading you did not send is not shown to the author. 9.4 Pausing ----------- Switch off "Offer me papers to read". You stay in the pool and receive no offers until you switch it back on. 9.5 Leaving ----------- Press "Leave the pool". You can rejoin later. Readings you have already sent stay with the papers they belong to (section 10). 9.6 What we do and do not keep ------------------------------ We keep no running count of how often you decline, lapse, run late, pause or leave, and nothing on the site depends on such a count. We do keep a short-lived record of each individual offer (currently for 30 days), so that the same paper is not offered to you twice, and then delete it. 9.7 If you need more time or a different arrangement ---------------------------------------------------- There is no timer on the reading form, and you can add time once yourself. If a disability or a circumstance means the usual arrangement does not work for you, write to accessibility@tillyard.org; the Accessibility Statement says more. You never need to explain a decision to decline. 10. What happens to your reading ================================ 10.1 Who sees it ---------------- The editor, and any paid staff who help run Tillyard Research, see your reading with your name. The author and any co-authors see your scores, your reasons, the strength and the first fix after the editor has decided, under a label such as "Reader B", without your name or college. They do not see your note for the editor. Other readers of the same paper do not see your reading. Nothing from it is shown publicly. 10.2 Your identity ------------------ We do not tell authors who read their papers, and the site is built so that they cannot find out from us. We could be required to disclose a reader's identity by a court order or other legal demand; the Privacy Notice says how we handle those. 10.3 How it is used ------------------- Your reading informs the editor's decision and gives the author feedback. Your scores are one of the things the editor weighs; they do not decide anything by themselves, and there is no threshold. If the paper is published, readers' scores are also used, by a published rule, to work out the term's Readers' Selection (see the Editorial Policy of Tillyard Research). The wording on a published paper counts how many readings the editor kept. 10.4 A reading can be set aside ------------------------------- The editor may set a reading aside, which leaves it out of the summary the editor works from, out of the count on the paper's page, out of Readers' Selection and out of your private count. Things that may lead to that include a reading returned within a few minutes, scores far from the other readers', text repeated from an earlier reading, or a conflict that came to light afterwards. Setting a reading aside is an editorial judgment about that reading. Nothing public says it happened, and we may not tell you. 10.5 It can be reported ----------------------- A reading is member text like any other. An author can report a comment, and a comment can be hidden while we look at it. 10.6 How long it is kept ------------------------ We keep readings for a period after the decision (currently 24 months), and offers that were declined or lapsed for a short time (currently 30 days). If you close your account, readings you wrote keep their scores and text until that period ends, with your identity removed from them. The Privacy Notice governs retention and says how to ask for your data. 10.7 Your rights in your comments --------------------------------- Your comments are your own writing. You give us permission to store them, to screen them, to show them to the editor and to the paper's authors as described above, and to keep them with the paper's records for the period in the Privacy Notice. We do not publish them. The author may use your comments to improve their paper. 10.8 Your reading count on The Record ------------------------------------- The private count described in 1.4 rises with each reading you send that the editor has not set aside. It is a statement of fact from Tillyard's own log. It says nothing about how well you read, and it is not a credential. 11. Concerns and misconduct =========================== 11.1 If you see a problem in a paper ------------------------------------ Use the note for the editor. Do not investigate it yourself, do not accuse the author, and do not take it anywhere else. If the paper suggests that someone is in immediate danger, use the report control so that it is seen quickly, and contact emergency services if that is what the situation calls for. 11.2 If someone asks you about a reading ---------------------------------------- If anyone contacts you claiming to be an author, or asks whether you read a particular paper, do not confirm or deny it. Tell us. 11.3 If you are an author and think a reader broke these guidelines ------------------------------------------------------------------- Use the report control on the reading, or the dispute form at tillyard.org/dispute. 11.4 What we may do ------------------- If we believe a reader has broken these guidelines, the editor may set their readings aside, and we may close the pool to that member through the ordinary case process described in the Community Guidelines and the Reporting, Enforcement and Appeals Policy. For serious breaches, such as sharing a paper or using an author's work, we may take the other steps those documents and the Terms of Use allow. The appeal described there applies. Any such step is about conduct. It is never about how much or how little someone reads. 11.5 Your responsibility ------------------------ A breach of confidence can do an author real harm. You are responsible for keeping to these guidelines, and for the consequences if you choose not to. Changes to this document ======================== We may revise these guidelines. When we do, we change the version number and the effective date at the top. If a change is significant, we may ask members in the pool to agree to the new version before they are offered another paper. The version you agreed to is recorded. Contact ======= - Questions about reading, or to tell the editor something about a reading you sent: support@tillyard.org - Accessibility: accessibility@tillyard.org - To dispute or correct something, with or without an account: tillyard.org/dispute - Legal notices: legal@tillyard.org, or by post to Tillyard, 10602 York Rd SW, Etna, Ohio 43062