Editorial Policy of Tillyard Research ===================================== Tillyard. Version 1.2. In effect from October 21, 2026. Fingerprint of this text (SHA-256): 6b465b18b481b34687d1b0ad20dd6dbb9df10f5fe411a23d08ba05c5ae6bc160 This version: https://tillyard.org/policies/editorial-policy/v/1.2 This policy is for authors, for readers in the reading pool, and for anyone who meets a paper from Tillyard Research on a résumé or in a search result and wants to know what stands behind it. - Tillyard Research is a student publication run by Tillyard. It is not an academic journal, and its readers are students rather than scholars in the field. - Authors choose their own topics and their own timing. Nothing is assigned, there is no fee, and nothing about how a paper is handled can be bought or hurried. - A paper may be read by up to three students at other colleges who chose to read, and who do not know who wrote it. Then one editor, the president, decides: approved, revise and resubmit (once), or not accepted for publication. - A published paper's page says exactly who read it and who approved it. Readers' scores are not published and there is no score a paper has to reach. - An author can appeal a decision once, with new information. There is one editor, so the second look is by the same person. - A published paper is not changed or removed silently. Corrections, withdrawals and retractions appear as dated notices, and a removed paper leaves a short notice page at its address. - Readers' Selection is worked out each term from readers' scores by a published rule, and confirmed by the editor. There is no prize. - Where there is a concern about copying, invented sources or undeclared AI help, we look at the sources, the drafting record and the author's answers to written questions. We act on editorial judgment. We make no public accusation and we do not claim to be able to detect AI-written text. - Tips, of any size, and sponsorships have no say in any editorial decision. If this summary and the full text below ever differ, the full text governs. 1. What Tillyard Research is =========================== 1.1 A student publication ------------------------- Tillyard Research is where members of Tillyard publish papers on topics they chose themselves. Its authors are college students. Its readers are other college students who choose to read. Its editor is one person. 1.2 Who runs it --------------- Tillyard Research is run by Tillyard (we, us), which is an Ohio nonprofit corporation. The editor is its president, and is named on the About page of Tillyard Research. The editor is one person and is not a specialist in most of the fields that papers cover. 1.3 What "published in Tillyard Research" means ---------------------------------------------- It means that a member wrote a paper at the desk on our site, declared how it was made, and sent it in; that it went through Tillyard checks in section 5; that it was read, where readers were available, by students at other colleges who chose to read it and who scored it against a published six-item rubric; and that the editor read it and approved it for publication. The paper's own page says how many readers there were. That is all it means. 1.4 What it gives the author ---------------------------- A permanent address for the paper, a page that says who read and approved it, and a Published Author item on the author's Record that anyone the author gives the link to can check. The author keeps the copyright; the Author Agreement and Licence sets out the terms. 1.5 Counts we publish --------------------- The About page of Tillyard Research shows, for each term, how many papers were submitted, published, not accepted and withdrawn, so that anyone can see how selective it has been. 1.6 How this policy fits with our other documents ------------------------------------------------- The Author Agreement and Licence governs the author's rights in a paper, the licence, the author's promises, and what remains after a paper comes down. The Reader Guidelines govern what readers must and must not do. The AI-Use Policy governs AI tools, the declaration and the drafting record. The Copyright Policy governs copyright complaints. The Community Guidelines and the Terms of Use govern conduct and accounts. The Privacy Notice governs what we keep. Where two documents cover the same point, the more specific one governs. 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. On how papers are handled and decided, that is this policy. 1.7 Numbers that can change --------------------------- Where this policy gives a number and says "currently", that number is a setting we can change. The number shown on the page when you act is the one that applies. 2. What Tillyard Research is =========================== 2.1 A student publication ------------------------- Tillyard Research is not a scholarly or academic journal and is not a journal of record. It is not peer-reviewed in the sense that universities and journals use those words: its readers are students who chose to read rather than scholars in the paper's field, and we do not check their expertise. 2.2 Credit and degrees ---------------------- Tillyard is not a school, a college or a course of study, and nobody accredits it. Publication here carries no academic credit, and we award no degree, certificate or mark. 2.3 Colleges ------------ We are not affiliated with or endorsed by any college or university. A college's name beside an author's name says where the author's email address was checked. It does not mean the college had any part in the paper. 2.4 Authors and readers ----------------------- Authors and readers are not employees, interns, contractors or volunteers of Tillyard. Writing or reading for Tillyard Research is not work experience and should not be listed as such. 2.5 What approval means ----------------------- The editor's approval is a decision to publish. It is not a statement that a paper's findings are correct, that anyone has repeated its analysis, or that we agree with it. 2.6 Professional advice ----------------------- Papers are student work. A paper that touches on law, tax, medicine, engineering or investments is not advice, and nobody should act on it as if it were. We may add a fixed label to such papers saying so. 2.7 Indexing and preservation ----------------------------- We issue no DOI for papers. We give search engines and scholarly indexes the information they use to list a paper, but whether any of them lists it is their decision. We have no arrangement with a preservation archive: if the site goes offline, the papers may go offline with it. 2.8 How to describe a paper here -------------------------------- List it under "Publications (student publication)" or "Writing", in the wording the author's Record gives. Do not call it peer-reviewed, refereed or a journal article. 3. What we publish ================== 3.1 Seven kinds of paper ------------------------ An author picks one kind and starts from its outline. The lengths and source counts below are current settings; the desk shows the ones that apply. Seven kinds of paper. Columns: kind, length in words, fewest sources, the two rubric items that matter most. Kind | Length in words | Fewest sources | The two rubric items that matter most ------------------------------------------------------------------------------ Research report | 1,500 to 6,000 | 6 | Method; the writer's own analysis Literature review | 1,500 to 5,000 | 12 | Sources; the writer's own analysis Case study | 1,200 to 4,000 | 5 | Evidence; limitations Data note | 600 to 2,000 | 2 | Method; evidence Policy memo | 800 to 2,500 | 5 | Question; the writer's own analysis Replication or fact-check | 800 to 3,000 | 2 | Evidence; method Profile, from public sources | 1,000 to 3,500 | 4 | Sources; limitations 3.2 Text, and figures --------------------- Papers are text. They can include headings, lists and quotations, and figures: pictures such as charts and diagrams, each with a caption and a description for readers who cannot see it. A paper can have 12 figures. Each is made smaller in the author's browser and is checked again by the server; it is shown only to the author, the readers and staff until the paper is published. Papers cannot include attached files or other uploads. Links in a paper are shown as text, except links to a DOI. If your work has data or code elsewhere, describe it in the data note and give its address as text, with an unchanging reference where the host offers one. 3.3 Topics are the author's --------------------------- We do not assign topics, set themes with closing dates, or commission papers. Work that is hard to fake is especially welcome: original data, a survey, interviews the author chose to do, code, a prototype, a replication. No kind of paper requires an interview or any private contact with anyone. 3.4 Language ------------ Papers are written in English. 3.5 What we do not publish -------------------------- The editor does not publish a paper that: 1. is not the author's own work, or was published elsewhere (the Author Agreement and Licence says what that means); 2. contains personal details of a private individual who has not agreed in writing; 3. breaks the Community Guidelines; 4. gives advice on law, tax, medicine, engineering or investments to a particular real person; 5. gives instructions meant to help someone cause serious harm; 6. is an advertisement or promotion for a product, a company, a candidate or an organization, including the author's own; 7. reproduces course materials that belong to an instructor, or confidential material that belongs to a lab, a university or an employer; 8. the editor cannot responsibly assess, for example because checking it would need specialist knowledge that neither the readers nor the editor has. 3.6 Writing about real people and organizations ----------------------------------------------- Profiles, case studies and fact-checks often discuss real people. We ask authors to hold to these standards: 1. rely on public sources and cite them, so that a reader can check each statement of fact; 2. keep private individuals out of it, unless they agreed in writing; 3. state allegations as allegations, say who made them, and rely on strong public sources for any claim of wrongdoing, such as court records, official findings or established news reporting. Do not accuse anyone of a crime or of dishonesty on your own inference; 4. quote accurately, and present the strongest version of a claim you are testing; 5. keep opinion and fact visibly apart; 6. write nothing that could identify a person under 18. The editor may ask an author whether they sought a response from a person or organization the paper criticizes, and may add an editor's note recording a response received later. 3.7 Research with people ------------------------ If a paper draws on interviews, a survey, observation or private information about identifiable people, the author says so on the submission form and describes in the data note how the information was gathered and what consent or review there was. The author is responsible for following their own college's rules for research with people. We are not an ethics board and do not review research designs. The paper's page shows what the author declared. 3.8 Declared interests ---------------------- Authors tell us about anything a fair reader would want to know when weighing the paper: who paid for or supported the work, a stake in its subject, a close tie to a person or organization it discusses. The editor sees the statement and may show it on the paper's page. 4. Who can submit, and the limits ================================= 4.1 Who ------- Any member in good standing can write at the desk. A new account can draft straight away and can submit once it is old enough (currently seven days). 4.2 The desk and its notice --------------------------- Papers are written at the desk on the site, which keeps a history of drafts and counts of typing and pasting. The first time you open the desk you are shown what it keeps and asked to agree. The AI-Use Policy describes this in full. 4.3 One paper in review at a time --------------------------------- A member can have one paper in review at a time, whether as lead author or as co-author. There is also a limit on how many papers a member can submit in a term (currently two) and on how many drafts a member can hold (currently three). These limits keep the line short enough for one editor. 4.4 Co-authors -------------- A paper can currently have up to three co-authors beside the lead author, chosen from members the lead author is already connected with. Each must agree to be listed and must approve the final text. See the Author Agreement and Licence. 4.5 No fee and no deadline -------------------------- There is no charge at any stage. No author has a deadline. The time windows mentioned in this policy are housekeeping: when one runs out, the paper goes back to the author's desk as a draft, and nothing is held against anyone. 4.6 Old drafts -------------- A draft that has not been touched for a long time (currently 365 days) is deleted, after a notice sent beforehand (currently at day 335). 5. From submission to decision ============================== 5.1 The steps at a glance ------------------------- The steps at a glance. Columns: stage, what happens, who or what acts, if time runs out. Stage | What happens | Who or what acts | If time runs out ---------------------------------------------------------- Writing | The author drafts at the desk, adds sources, and can run "Check my paper" | The author | An untouched draft is eventually deleted (4.6) Submitting | Four steps: Check, Declare, Licence, Send. The text is frozen | The author | Nothing; there is no deadline House checks | Text-safety screening, DOI look-ups, and a comparison of sections | Automated | The paper keeps its order and waits Co-author approval | Each co-author reads the frozen text and approves it | Co-authors | Back to draft (currently after 14 days) The line | The paper takes a place in a capped line | Automated | It waits in order for a place Readers | The paper is offered to up to three readers | Members of the pool, if they choose | Offers lapse and unreturned readings are released to other readers To the editor | When three readings are in, when the wait ends, or when nobody has been free to read | Automated; the editor can also call a paper up early | Not applicable Questions (optional) | The editor may ask up to five written questions | The editor, then the author | Back to draft (currently after 14 days) Decision | Approved, revise and resubmit, or not accepted | The editor | We give no time for a decision Revision (if asked) | The author revises and resubmits once | The author | Back to draft (currently after 180 days) Appeal (if wanted) | One appeal with new information | The author, then the editor | The window closes (currently 14 days after the letter) 5.2 Declarations ---------------- At the Declare step the author confirms that the paper is their own work and that they cite other people's words and ideas; that every author listed has read this version; that it names no private person's personal details unless that person agreed in writing; and that it has not been published elsewhere. The author also states where the work came from (on their own; for a class; in a lab or research group; for an employer or client), what part people play in it (none; public sources only; interviews; a survey), and what AI help they had, with a note. Where the work came from a lab or an employer, the form warns that someone else may own it. At the Licence step each author accepts the Author Agreement and Licence. 5.3 Tillyard checks ------------------ When the author presses Send, the text is frozen and split into sections. Three checks then run without anyone at Tillyard reading the paper: 1. Text-safety screening. The same automated screening that applies to other member text. If something is held, the paper stays visible only to its authors and a person at Tillyard looks at it. 2. DOI look-ups. Each DOI in the source list is looked up in Crossref's public index. The result is recorded as found, not found, or found with details that differ from what the author typed. 3. Identical sections. The frozen sections are compared with papers already in the library and with the same author's earlier papers that were not accepted. This finds sections that are word-for-word identical. It is not a search of the wider web, and it is not a plagiarism service. These checks report facts to the editor. They produce no mark, and a paper does not succeed or fail at them. One exception is automatic: if more than half the sections of a new submission are identical to the same author's paper that was not accepted in the recent past (currently 180 days), the submission is refused and the author is pointed to the appeal in section 9. 5.4 The queue for review ------------------------ The number of papers in review at once is capped. A paper's status page shows its place in line and an estimate of the wait, which is an estimate and not a promise. If the line is full, the paper is ready and waits in order for a place; the author can ask to be told when one opens. 5.5 Readers ----------- A paper in line is offered to readers from the open pool, up to three. Each reader is drawn at random and may decline. The draw leaves out anyone who is an author, anyone at an author's college, anyone connected with or blocked by an author on the site, and anyone who has recently read these authors or read alongside the paper's other readers. For the first few days only readers who listed the paper's field are drawn (currently three days); after that, any reader. Readers see the paper without the authors' names or colleges. Each scores six rubric items from 0 to 2 with a reason for each, names one strength and the first thing to fix, and checks two of the paper's sources, typing what they found. Readers do not see one another's readings. The Reader Guidelines set out their duties, including confidentiality. 5.6 When the paper goes to the editor ------------------------------------- A paper goes to the editor at the earliest of three moments: when three readings are in; when the wait ends (currently 14 days after it took its place); or when nobody in the pool has been free to read it for a time (currently two days). The editor can also call a paper up early. Readings still out can arrive until the editor decides. So a paper may be decided with three readings, with two, with one or with none, and the wording on its page reflects that (section 7). 5.7 What the editor has in front of them ---------------------------------------- The editor reads the paper and a one-screen brief. The brief holds the readers' scores side by side and how far apart they are; each reader's reasons and source checks; notes on any reading that looks unreliable; the results of the DOI look-ups and the section comparison; the author's declarations; the drafting record, as background; and six distinctive sentences picked from the paper, which the editor may search for on the web by hand. 5.8 Setting a reading aside --------------------------- The editor may set a reading aside, for example where it was returned within a few minutes, where its scores sit far from the others', where its text repeats an earlier reading, or where a conflict has come to light. A reading set aside is left out of the editor's summary, out of the count on the paper's page, and out of Readers' Selection. 5.9 Questions ------------- The editor may ask the author up to five written questions about the paper. They are asked and answered in a form on the site, with no timer. The AI-Use Policy describes the kinds of question and what the editor will not ask. 5.10 The decision letter ------------------------ The editor decides and sends a letter. Letters begin from a house template and are finished by the editor. The letter carries one of three headings: "Approved for publication", "Revise and resubmit", or "Not accepted for publication". With it the author sees each reading, labeled Reader A, B or C, with its scores, reasons, strength and first fix. The author can press "This helped" on a reading, which sends that reader a thanks with no names. There is no channel for replying to a reader. 5.11 Publication ---------------- An approved paper is published by the editor, which may be at once. Publication creates the paper's public page, a Published Author item on each author's Record, and listings in the library. 5.12 Withdrawing ---------------- The lead author can withdraw a paper at any time before publication. After publication, any author can. See section 11 and the Author Agreement and Licence. 5.13 What we do not promise --------------------------- We do not promise that a paper will be read by any number of readers, that it will be decided by any date, or that it will be published. 6. How the editor decides ========================= 6.1 What the editor weighs -------------------------- The six rubric items, with most attention to the two that matter most for the kind of paper; the readers' scores and reasons; the readers' source checks and the DOI results; the author's declarations; the drafting record and any answers to questions, as background; whether the paper fits what Tillyard Research publishes (section 3); and any legal or safety concern. 6.2 Readers inform; the editor decides -------------------------------------- Readers' scores are advice. There is no score that earns publication and none that rules it out. A paper with high scores can be declined, and a paper with mixed scores can be approved. No reader recommends an outcome. 6.3 The editor's discretion --------------------------- The editor may decline any paper, for reasons of quality or for reasons that have nothing to do with quality, such as legal risk, a subject the editor cannot responsibly assess, or the capacity of one person. The editor is not obliged to give reasons beyond the letter. 6.4 What the editor does not weigh ---------------------------------- Readers do not know who wrote a paper. The editor does. The editor does not decline a paper because of who its author is: their race, color, religion, sex, sexual orientation, gender identity, national origin, disability, age or college. Whether a member has ever left a tip, or has not, plays no part (section 15). 6.5 What approval does not cover -------------------------------- The editor does not re-run analyses, open every source, verify data or test code. Responsibility for what a paper says stays with its authors. 6.6 Nobody else decides, and nothing decides automatically ---------------------------------------------------------- No reader, sponsor, payer, college or employer has a say in a decision. No paper is approved by an automatic rule: every published paper was approved by the editor in person, so that the words "approved by the editor" stay true. 6.7 If others are added later ----------------------------- If a second paid editor or a named faculty reader ever joins Tillyard Research, this policy would be revised first, and the wording on a paper's page would say who approved it. 7. What the wording on a paper's page means =========================================== 7.1 The line under "Who read this paper" ---------------------------------------- Every published paper carries one of these lines. Which one depends only on how many readings the editor kept. The line under "Who read this paper". Columns: readings the editor kept, the line on the page. Readings the editor kept | The line on the page ----------------------------------------------- Three | "Read by three students at other colleges and approved by the editor" Two | "Read by two students at other colleges and approved by the editor" One or none | "Approved by the editor" A paper that was revised before approval carries a second line: "The readers read the first version. The editor approved the revision." 7.2 What "read by ... students at other colleges" means ------------------------------------------------------- Each of those readers: 1. was a member of Tillyard who controlled an email address at a college different from every author's college, checked by us; 2. was not connected with any author on the site, and was drawn at random; 3. did not know who wrote the paper; 4. scored the six rubric items, gave a reason for each, opened two of the paper's sources and typed what they found; 5. sent a completed reading that the editor did not set aside. 7.3 What it does not mean ------------------------- It does not mean that the readers were experts in the subject, were faculty, or had any qualification; we check a reader's college email address and nothing else about them. It does not mean that the readers recommended publication, agreed with one another, or agreed with the paper. It does not mean that every source or any of the data was checked. 7.4 What "approved by the editor" means --------------------------------------- The editor of Tillyard Research read the paper and the brief described in 5.7 and decided to publish it. Section 2.5 says what that does not mean. 7.5 The other lines ------------------- The same panel may show: 1. the date of approval and the date of publication; 2. "Readers scored it against the published rubric", with a link to the rubric; 3. "AI help, as the author declared it", followed by the author's own declaration and note. This is the author's statement. It is not a finding by us; 4. where the work came from, as the author declared it (for example "Made on the author's own initiative"); 5. what the author declared about people in the paper, and the data note; 6. how many sources are listed and how many of the paper's DOIs were found in Crossref's public index on a stated date. A DOI being found means a record with that identifier exists. It does not mean the source supports what the paper says. 7.6 Scores stay private ----------------------- Readers' scores are shown to the author and the editor only. They are never shown on a paper's page, in a list or on The Record, and they are never described as a mark. 7.7 The Record -------------- Each author's Record item for the paper repeats the line in 7.1 and adds that the paper was an unpaid, self-directed contribution and that the author was not an employee, intern or contractor of Tillyard. That record page shows its status on the day it is opened. 8. Revise and resubmit ====================== 8.1 Once -------- The editor may ask for a revision once. There is no second round. 8.2 How revising and resubmitting works --------------------------------------- The desk reopens for that paper. The author revises, writes a short note headed "What I changed", and resubmits. If there are co-authors, the lead author resubmits only a version that all of them have agreed to. Tillyard checks run again on what changed. 8.3 Who looks at the revision ----------------------------- The revision goes to the editor, who can compare it with the first version. Today it does not go back to the readers. That is why a revised paper's page says: "The readers read the first version. The editor approved the revision." 8.4 The outcome --------------- The editor approves the revision or does not accept it. 8.5 Time -------- The author has a long window to revise (currently 180 days). If it runs out, the paper goes back to the desk as an ordinary draft. 9. Not accepted, and appeals ============================ 9.1 What the author receives ---------------------------- A letter headed "Not accepted for publication", the editor's notes against the rubric, and the readers' scores and comments. The letter makes no accusation. 9.2 What is public ------------------ Nothing. A paper that was not accepted has no public page, appears in no list, and leaves no mark on the author's Record. It is counted, without a name, in the totals in 1.5. We keep the manuscript and its records privately for the period in the Privacy Notice (currently 24 months). 9.3 What the author can do with the paper ----------------------------------------- It is theirs. They can publish it anywhere else. They can rework it and submit it here as a new paper, subject to the rule in 5.3 about a submission that is mostly identical to the one not accepted. 9.4 One appeal -------------- An author can ask once for a second look, within a set time after the letter (currently 14 days). The appeal is a short written statement of new information: what the editor should know that was not in the paper or the readings. Examples: a reading misread a section; the letter contains a factual mistake; a source check failed only because of a paywall; a reader appears to have had a conflict. Disagreeing with the outcome is not, by itself, new information. 9.5 Who looks again ------------------- The same editor, after a short wait (currently at least three days). There is one editor, so we cannot offer a second person. 9.6 The outcome --------------- The editor replies in writing. The decision stands, or the paper is invited back for revision, or it is approved. That reply is final so far as Tillyard Research is concerned. 9.7 No consequence for appealing -------------------------------- Appealing does not affect an author's standing or any later paper. 9.8 Other decisions ------------------- A decision about a correction, a retraction, the reading pool or an account is contested through the dispute form at tillyard.org/dispute (see the Disputes and Corrections Policy) or the appeal described in the Reporting, Enforcement and Appeals Policy, not through this section. 10. Conflicts of interest ========================= 10.1 Authors ------------ See 3.8. An undeclared interest that comes to light later may lead to an editor's note or a retraction. 10.2 Readers ------------ See the Reader Guidelines. A reader who knows or can guess the author, or who has something to gain, steps aside. 10.3 The editor's standing interest ----------------------------------- The editor founded Tillyard, is its president and runs it day to day. We state that here so that nobody has to infer it. 10.4 The editor and particular authors -------------------------------------- The editor sees authors' names. Where the editor has a close personal tie to an author (a relative, a partner, a close friend, a current collaborator, or a business relationship), the editor may decline to handle the paper, and tells the author why. While there is only one editor, that means the paper cannot be considered here. Where the tie is slighter (the same college, a shared class, an acquaintance), the editor records it privately with the decision and decides. 10.5 The editor's own writing ----------------------------- The editor does not submit papers to Tillyard Research. 10.6 Papers about Tillyard ------------------------- The editor may decline a paper whose subject is Tillyard, its owner, the owner's other business, or anyone who has sponsored or made a larger tip to Tillyard, because the editor could not decide it without an interest. 10.7 Staff ---------- If paid staff later help run Tillyard Research, the same rules apply to them. 11. After publication: corrections, editor's notes, withdrawals and retractions ============================================================================== 11.1 A published paper is changed in the open --------------------------------------------- A published paper is part of a public record that other people may have cited, linked to or checked. So changes to it are dated and visible, and a paper that comes down leaves a notice at its address. The one temporary exception is a hold (11.7). 11.2 Corrections ---------------- An author can ask for a correction from the paper's page (currently up to three requests for a paper in a year). Anyone else can point out an error through the report control or the dispute form. The editor decides. A correction fixes a mistake: a wrong figure, a misquotation, a wrong citation, a misattribution, a misspelled name. It is not a way to rewrite a paper or add new work. When a correction is made: 1. the text is changed; 2. a dated notice above the title says what was wrong and what the text now says; 3. the earlier version is kept in our private records; 4. the Record items for the paper are unchanged. The editor may refuse a request, and we aim to tell the person who asked. 11.3 Editor's notes ------------------- The editor may place a dated note above a paper's title: to point to a later or fuller version elsewhere, to record a response from a person or organization the paper discusses, to disclose an interest that should have been declared, or to say that a question about the paper is being looked at. 11.4 Withdrawal by an author ---------------------------- Any author of a published paper can withdraw it from the paper's page. The paper leaves the library at once, and when the editor has confirmed the request the page becomes a tombstone reading "Withdrawn at an author's request on (date)." Withdrawal is the author's right and carries no judgment. On a co-authored paper, one author's withdrawal takes the paper down for all of them. The Author Agreement and Licence sets out the details and governs. 11.5 Retraction by the editor ----------------------------- The editor may retract a published paper where: 1. it contains a serious error that a correction cannot repair, so that its main findings cannot be relied on; 2. a promise the author made about it turns out to be untrue (for example about authorship, sources, permissions or AI help); 3. it was published elsewhere first; 4. there is a legal or safety reason; 5. the editor can no longer stand behind having approved it. 11.6 How a retraction is handled -------------------------------- We aim to tell the authors what the concern is, with enough detail to answer it, and to hear from them before the editor decides. Where there is a legal or safety reason for urgency, we may hold the paper first. A retracted paper's page becomes a tombstone that carries one of a short set of fixed, neutral lines, such as "Retracted by the editor on April 2, 2027." No line alleges misconduct, and we publish no reason that does. The authors are told the reason privately. The Record items for the paper are withdrawn. 11.7 Holds ---------- We may hide a paper for a time while we look at a report, a dispute or a legal complaint. During a hold the paper's page answers as "not found", the paper is left out of every list, and each author's Record item for it reads "no longer listed". Nothing public says that a hold exists, because a hold is not a finding. It ends when the paper is restored, corrected, withdrawn or retracted. 11.8 Tombstones --------------- A tombstone stays at the paper's original address. It shows the title, the volume and number, the dated notice, and the words "This paper is no longer published here. The page stays so that links keep working." It shows no author, no college and none of the text, and it asks search engines not to list it. The paper's number in its volume is not reused. If the title itself contains someone's personal details, or the law requires it, we may shorten or remove the title. 11.9 What The Record shows -------------------------- When a paper is withdrawn or retracted, each author's Published Author item for it is withdrawn, and its record page reads "This item was withdrawn", with the month. That page shows no name, no title and no reason. Authors are told to take the paper off their résumés and profiles. A Readers' Selection for that paper is withdrawn with it. The table in the Author Agreement and Licence shows what each page displays after each kind of event. 11.10 What we cannot do ----------------------- We cannot recall copies made by other people, and we cannot make a search engine or a web archive drop its copy; that is up to them and can take days or weeks. 11.11 Questions from other people --------------------------------- If a college, an employer, a journalist or anyone else asks why a paper was withdrawn or retracted, we point them to the public notice. We do not discuss an author's case with other people unless the law requires it or the author asks us to. 12. Readers' Selection ====================== 12.1 What Readers' Selection is ------------------------------- Readers' Selection is not running at present. This section describes how it would work if we open it, and it will say so when we do. Each term, Readers' Selection lists papers published that term, chosen by the rule in 12.4 from the totals their readers gave. It is recorded on each selected author's Record beside the paper. It is not an honour or an award. There is no prize and no money, and nobody applies for it. 12.2 Terms and volumes ---------------------- Tillyard Research publishes in three terms a year, each forming one volume: January 1 to May 31, June 1 to August 14, and August 15 to December 31. 12.3 Which papers are eligible ------------------------------ A paper is eligible if it was published in that term, has at least two readings that the editor kept, and is not held, withdrawn or retracted when the list is worked out. A paper published with fewer than two kept readings is not eligible, because there is too little to compare. 12.4 How the list is worked out ------------------------------- Each reading gives a total out of 12 (six items, each 0 to 2). Eligible papers are put in order by: 1. the median of the paper's reading totals, highest first; 2. then the mean of those totals; 3. then the mean score on the two rubric items that matter most for that kind of paper (the table in 3.1); 4. then the earlier publication date. 12.5 How many are selected -------------------------- The top tenth of the eligible papers, rounded up, with at least one and at most five (currently). If fewer than five papers are eligible in a term, there is no Readers' Selection that term. 12.6 An example --------------- In a term with 23 eligible papers, a tenth is 2.3, so three papers are selected. A paper with reading totals of 9, 10 and 11 has a median of 10 and a mean of 10. A paper with totals of 8, 10 and 12 has the same median and the same mean, so the two are separated by step 3, and if still level by step 4. 12.7 When, and who confirms --------------------------- The list is worked out automatically a short time after the term ends (currently 14 days) and put to the editor. The editor confirms it. The editor may leave a paper out, for example because a correction or a concern about it is pending, but does not add a paper that the rule did not select and does not reorder the list. The selection is then recorded. 12.8 What is public ------------------- The list for each term and this rule. The scores are not published. A selected author's Record says, for example, "One of the 3 papers picked from the 23 read by students that term, confirmed by the editor." No public line mentions a score. 12.9 What it does and does not show ----------------------------------- Readers' Selection reports how student readers scored papers. Each paper had its own two or three readers, not a single panel that read them all, and the numbers are small. It is a fair summary of the readings and a rough measure of anything more. It is not a ranking by experts. 12.10 Later changes ------------------- If a selected paper is later withdrawn or retracted, its selection is withdrawn with it. The list is not worked out again and no other paper takes its place. 12.11 No lobbying ----------------- Nobody can ask for, arrange or pay for a selection. Section 13 covers attempts to interfere with readings. 13. Copying, invented sources and other misconduct ================================================== 13.1 What we mean ----------------- - Plagiarism is presenting another person's words, ideas, data or structure as your own without credit. It includes copying, close paraphrase without credit, translating someone's text without credit, and reusing your own published work without saying so. - Fabrication is making up data, results, quotations, interviews or sources. - Falsification is changing or leaving out data or results, or misdescribing a method, so that the paper gives a false picture. - Undeclared AI help is help from an AI tool that the AI-Use Policy requires to be declared and that was not, or that was misdescribed. - Authorship problems include listing someone who did not contribute, leaving out someone who did, and submitting without a co-author's agreement. - Duplicate publication is publishing here what was published elsewhere, or publishing the same paper here twice in different words. - Undeclared interests are those in 3.8. - Problems in research with people include gathering information without consent and publishing details that identify a private person. - Interfering with the reading includes arranging who reads a paper, using a second account, trying to identify or contact a reader, and offering anything to a reader or to the editor. - Misusing the process includes reporting a paper in bad faith to have it taken down, and asking for corrections that are not true. 13.2 How a concern reaches us ----------------------------- Through Tillyard checks; through a reader's note for the editor; through the editor's own reading and hand searches; through the report control on a paper or a reading; through the dispute form, which needs no account; or by a letter or email from someone outside. 13.3 What we look at -------------------- The paper and its sources; the DOI results; what readers typed when they opened the drawn sources; the section comparison; the author's declarations; the drafting record, as background; the author's answers to written questions; and whatever the person raising the concern has supplied. 13.4 What we do not rely on --------------------------- We review submissions and may decline or remove content we believe is AI-generated, plagiarized or inaccurate. This is editorial judgment; we do not guarantee detection. We do not use AI-detection services, and we do not act on a detector score, whoever produced it. We do not use an outside plagiarism service. A report that gives no particulars carries little weight. Typing and paste counts never decide anything by themselves. 13.5 Before publication ----------------------- The editor may ask the author written questions. If the editor is not satisfied that the paper is the author's own, accurate and properly sourced, the paper is not accepted. The letter says "Not accepted for publication" and gives notes against the rubric. It makes no accusation, and nothing is public. 13.6 After publication ---------------------- We may hold the paper where that is needed. We aim to put the concern to the authors and hear their answer. The editor may then leave the paper as it is, correct it, add an editor's note, or retract it (section 11). 13.7 Fairness ------------- We tell the author what the concern is in enough detail to answer it, without naming whoever raised it. We give the author a chance to answer. The editor's conclusion is an editorial judgment about whether Tillyard Research should carry the paper. It is not a finding of guilt, and we are not a court or a disciplinary body. An author who disagrees can use the appeal in section 9 for a paper not accepted, or the dispute form for anything else. 13.8 Honest mistakes -------------------- Most errors are honest ones. An author who finds and reports their own mistake will find us glad to correct it. A correction is not a mark against anyone. 13.9 The account ---------------- Whether anything happens to a member's account is a separate question, decided under the Community Guidelines, the Reporting, Enforcement and Appeals Policy and the Terms of Use, with the appeal described there. Deliberate or repeated misconduct may lead to a suspension or to removal. 13.10 What we do not do ----------------------- We do not publish accusations. We do not put reasons on tombstones. We do not make a practice of reporting authors to their colleges, their employers or anyone else. Serious misconduct such as threats, stalking, fraud or impersonation may lead to permanent removal and, where appropriate and lawful, referral to law enforcement or your institution. 13.11 People who raise concerns ------------------------------- We do not tell an author who raised a concern about their paper. A court order could require us to. Reports made in bad faith are handled under the Community Guidelines. 13.12 Readers ------------- A reader who shares a paper, uses what they read, hides a conflict or sends a reading they did not do has broken the Reader Guidelines, which say what may follow. 13.13 Disputes about credit --------------------------- Where authors disagree about who should be listed or who did what, we do not decide who is right. We may add a note, hold the paper, or take it down until they agree. 14. Takedown and removal requests ================================= 14.1 How to ask --------------- Anyone can ask us to look at a paper, with or without an account: 1. the report control at the foot of every paper; 2. the dispute form at tillyard.org/dispute, which gives a ticket number and a status page (the Disputes and Corrections Policy describes it); 3. for copyright, a notice to copyright@tillyard.org, as the Copyright Policy describes; 4. for legal papers, legal@tillyard.org or post to Tillyard, 10602 York Rd SW, Etna, Ohio 43062. 14.2 We look at the substance ----------------------------- The editor selects and approves every paper in Tillyard Research. So we do not treat a complaint about a paper as a formality to be processed. We look at whether it is right. 14.3 Kinds of request --------------------- 1. "This copies my work." Handled under the Copyright Policy, which sets out what a notice must contain, how the author can respond with a counter-notice, and how repeat infringement is treated. That policy governs. We also look at the complaint as a possible case of plagiarism under section 13, which does not depend on copyright. 2. "This paper is about me." A person named or described in a paper can tell us that it contains their personal details, or names them without their agreement. Authors promise that no private individual's personal details appear without written agreement. 3. "This is false and it harms me." Tell us which statements are false and, if you can, what shows it. 4. "This belongs to us." A class instructor, a lab, a university, an employer or a research sponsor can tell us that a paper uses material the author had no right to publish. 5. "I should be credited" or "I should not be." See 13.13. 6. "This is confidential." Tell us what is confidential and what makes it so. 7. A court order or a request from a public authority. Send it to legal@tillyard.org. We comply with valid legal orders. How we handle requests for information about members is in the Privacy Notice and the Law Enforcement and Legal Requests Policy. 8. Intimate images. Tillyard Research holds no images except the figures authors add to their papers, each of which the editor can see before a paper is published. The site has a separate removal process for intimate images, which applies everywhere on the site and is set out in the Intimate-Image Removal Policy. 14.4 What we may do ------------------- We may hold the paper while we look. We may ask the person complaining for particulars, and ask the author for an answer. The editor may then restore the paper, correct it, remove a passage and say so in a notice, add an editor's note, or retract it. We aim to tell the person who asked, and the authors, what was decided. We give no time for a decision beyond what a law requires. 14.5 The result is visible -------------------------- If a paper is changed, a notice says so. If it comes down, a tombstone stays (11.8). The exceptions are a hold while we look (11.7), and a case where a court order or the law requires that even the title be removed. 14.6 Requests from colleges and employers ----------------------------------------- A college, an employer or anyone else with a complaint about a paper is heard like any other complainant. None of them has authority over what Tillyard Research publishes, and none is given information about an author except as the Privacy Notice allows. 14.7 Abuse of the process ------------------------- A request made in bad faith, to silence fair criticism or to harm a rival, is handled under the Community Guidelines. Asking in good faith never costs a person their standing. 15. Money and editorial decisions ================================= 15.1 No fees ------------ Authors pay nothing to submit, to be read, to be decided, to be published, to be corrected or to withdraw. Readers are paid nothing. There is no paid tier and no way to pay for faster handling. 15.2 Tips --------- Tillyard cannot accept donations yet. When it can, anyone may leave one on a payment page run by Stripe. A donation is paid to Tillyard for its general work. It is not paid to an author and is not tied to a paper. Donating does not change anyone's standing, and it plays no part in any editorial decision. The site keeps no record linking a tip to a member. The president, who is also the editor, can see a payer's name, email address and amount in Stripe's own records, as the Privacy Notice says; that information is not used in handling papers. 15.3 Where the tip button does not appear ----------------------------------------- It does not appear on the desk, on the submission steps, on the reading form, on a decision letter or on an appeal, so that it cannot be read as pressure or as payment for an outcome. 15.4 Larger payments and sponsorships ------------------------------------- These are arranged in writing. They give no ownership, no return, and no say over editorial decisions, readings or anyone's standing. The Tips and Larger Payments Terms govern tips, refunds and larger tips. 15.5 No paid or sponsored papers -------------------------------- We do not publish a paper in return for payment, and we do not publish sponsored content or advertising in Tillyard Research. 15.6 If a sponsor is ever named in Tillyard Research --------------------------------------------------- If a sponsor ever supports something in Tillyard Research, such as an open call on a theme, the page will say so. A sponsor does not choose, read or score papers, does not see unpublished manuscripts, scores or readers' identities, and is not told the outcome of any paper before it is public. An author whose paper concerns a sponsor declares that (3.8), and the editor may decline to handle it (10.6). 15.7 Nothing of value to the editor ----------------------------------- The editor does not accept anything of value from an author or a reader in connection with a paper. 15.8 Employers -------------- No employer or recruiter pays for access to papers, authors or readers. If that ever changes, this policy will be revised before it does. 16. How Tillyard uses automated tools ==================================== 16.1 Screening -------------- Text sent to Tillyard Research passes through automated text-safety screening, which may use outside services named in the Privacy Notice. Screening can hold text for a person to look at. It does not decide whether a paper is published. 16.2 No automated decisions about papers ---------------------------------------- No paper is scored, approved or declined by software. The AI-Use Policy says more, including what Tillyard does and does not do with AI tools. 17. Records =========== 17.1 What we keep ----------------- The Privacy Notice governs. In outline, and as current settings: manuscripts that were withdrawn or not accepted, with their readings and decisions, are kept privately for 24 months; draft history is thinned when a paper is decided and deleted 90 days later, except that the published version is kept while the paper is published; readings are kept for 24 months after the decision; a published paper and its notices are kept while it is published; a tombstone keeps the title, the volume and number, and the notice. 17.2 The staff log ------------------ Every action the editor or staff take on a paper is written to a log that cannot be edited, including by the owner. 17.3 Your own copy ------------------ Members can export their own submissions, the readings they received (without readers' identities) and the readings they wrote. The Privacy Notice explains how. Changes to this document ======================== We may revise this policy. When we do, we change the version number and the effective date at the top. A paper is handled under the policy in force when each step is taken. Where a change is significant, we aim to tell members through the site before it takes effect. Contact ======= - Questions about Tillyard Research or about a paper: support@tillyard.org - To dispute or correct something, with or without an account: tillyard.org/dispute - Copyright complaints: copyright@tillyard.org, as the Copyright Policy explains - Legal notices and court orders: legal@tillyard.org, or by post to Tillyard, 10602 York Rd SW, Etna, Ohio 43062 - Accessibility: accessibility@tillyard.org