Sentences you can send, not principles you have to translate.

Academic boundary scripts: what to say to a supervisor, a co-author, an editor and a collaborator

Sixteen situations, each with the words. Written down because the moment you need this is the moment you have least appetite for working out how it is supposed to go.

No ghostwriting, ever · Human-led, PhD-signed

The page is the content. The download is the working copy — one situation per printed page, so you can take the single page you need into a supervision without carrying the other fifteen. One email covers all six Vault resources.

1. 01 · A supervisor asks for a chapter over the holidays

   What is actually being asked

   Usually not for a chapter. Usually for reassurance that the thesis is moving, sent in the form of a deadline.

   The position

   Give the reassurance, on a date you choose.

   I’m not going to have the chapter for [DATE]. I’m taking [DATES] off, and I’ll send you [SPECIFIC SMALLER THING: the revised outline, the results section, the analysis plan] on [DATE AFTER YOU RETURN]. If the timing is driven by something at your end that I don’t know about, tell me and we’ll work out what actually has to move.

   Thank you for the push on this. I want to be straight with you about the schedule rather than send you something half-finished on [DATE]. I’m away [DATES]. What I can commit to is [SPECIFIC DELIVERABLE] by [DATE], and the full chapter by [DATE]. If that creates a problem for [PROGRESSION MILESTONE / FUNDING DATE], let me know this week and I’ll rebuild the plan around it.

   Do not send. A list of everything else you are doing. It reads as a bid for sympathy and it invites a conversation about your priorities rather than about the date.

2. 02 · A collaborator adds themselves as an author

   What is actually being asked

   Nothing. It has already been done, which is why this one needs a written record and not a conversation in a corridor.

   The position

   Ask what the contribution was, in writing, referring to the criteria rather than to the person.

   I’ve just seen [NAME] on the author list on the current draft. I don’t have a record of their contribution, so I want to get this right before we submit. Can you set out what they contributed against the journal’s authorship criteria — the design, the data, the analysis or the drafting? If it’s an acknowledgement rather than authorship, let’s put them there properly. I’d rather sort this now than have it raised after publication.

   Before I finalise the submission I want to confirm the author list, since I’m the one signing the authorship statement. [JOURNAL] requires each author to have contributed to [CRITERIA] and to approve the final version. Could you tell me which of those applies for [NAME], so I can complete the contribution statement accurately? I’m asking because the statement carries my name against it.

   Why that last sentence works. It moves the question from a challenge to a senior colleague to a compliance requirement you personally sign. That is also true, which is why it holds.

   If it is refused or ignored. Stop negotiating and contact your research integrity office. Authorship disputes have a process precisely because they cannot be settled between two people with unequal standing. Keep the emails.

3. 03 · Declining a peer review

   What is actually being asked

   Two to eight hours of unpaid expert work, on a schedule set by someone else.

   The position

   Decline in three lines, and name a successor if you have one. Editors remember a fast no far more warmly than a slow yes.

   Thank you for the invitation. I’m not able to take this one on by [DATE]. [ONE REASON: my review load is full this quarter / this sits outside my area on [SPECIFIC ASPECT].] [NAME] at [INSTITUTION] works directly on [TOPIC] and may be a better fit.

   I can’t do this one by [DATE], but I can take a review in [MONTH] if that is useful to you.

   Do not send. A paragraph explaining how busy you are. Every editor is also busy and it is not a reason they can act on.

4. 04 · Asked to review outside your expertise

   What is actually being asked

   For your name to stand behind a judgement you are not equipped to make.

   The position

   Name the exact limit rather than declining generally, because a partial competence is useful information for an editor.

   I can review the [PART YOU CAN JUDGE: the design, the statistics, the framing] but not the [PART YOU CANNOT: the assay, the archival claims, the modelling]. If a partial review on that basis is useful, I’ll do it and I’ll say in the report exactly what I did not assess. If you need a full review, I’d decline this one.

   Why this is worth sending rather than a plain no. A review with its own boundary stated is more useful to an editor than a confident review of a section the reviewer could not evaluate. It is also the version you can defend later.

5. 05 · A co-author has gone silent and the deadline is here

   What is actually being asked

   For you to absorb the risk of somebody else’s silence.

   The position

   One clear message with a decision point and a date, then act on the date.

   I haven’t heard back on the draft I sent on [DATE], and the submission window closes on [DATE]. If I don’t hear from you by [DATE, at least 5 working days out], I’ll submit the version attached, with your name on it and your contribution as recorded in the statement. Tell me if you’d rather I hold it. If you want out of the paper, that’s fine too — say so and I’ll adjust the author list.

   The escalation, if the silence continues past the date. Send the third option in writing and copy nobody. Then submit, or withdraw them, and keep the record. A named deadline that you then ignore teaches everybody that your deadlines are decorative.

6. 06 · Asked to add citations you do not think belong

   What is actually being asked

   For your reference list to serve somebody’s metrics.

   The position

   Agree to consider anything relevant, decline the rest on a stated ground, once.

   Thank you for the suggested references. I’ve read [N] of them and added [THE ONES THAT ARE RELEVANT] at [LOCATION], where they support [SPECIFIC POINT]. I haven’t added the remainder. They address [TOPIC], which the paper does not make a claim about, and adding them would misrepresent what the argument rests on. If there’s a specific point where you think one of them is load-bearing, tell me which and I’ll look again.

   I’ve added [THE RELEVANT ONES]. The others I’ve left out because the paper doesn’t make a claim in that area, and I’d rather the reference list matched the argument. Happy to revisit if you can point me at the specific paragraph.

   The line underneath all of this. Citations follow the argument. A reference list assembled for any other reason is a data-integrity problem, and journals increasingly treat it as one.

7. 07 · Asked to take on teaching or admin you have no room for

   What is actually being asked

   For you to solve a resourcing problem privately, by absorbing it.

   The position

   Make the trade visible. Never refuse in the abstract, always refuse against a list.

   I can take [THE THING], but not alongside [EXISTING COMMITMENT] and [EXISTING COMMITMENT] in the same period. If it’s the priority, tell me which of those two moves and I’ll pick it up. If neither can move, I’d need this to go to somebody else this time.

   Yes, I can do it. So that I deliver it properly, I need to know what comes off the list — my current commitments this term are [1], [2] and [3], with [DEADLINE] on [DATE]. Which of those would you like me to deprioritise?

   Why this works when a plain no does not. It converts your capacity from an attitude into a scheduling question, which is the other person’s job to answer. Most requests of this kind are made without a picture of your load, and the picture is often enough.

8. 08 · Asked to write something that goes out under someone else’s name

   What is actually being asked

   For you to take a professional risk that will not be recorded anywhere.

   The position

   Decline the writing, offer the reading. This is the line the whole of academic integrity policy is built around, and it does not bend for seniority or for friendship.

   I’m not able to write [THE THING] for you to submit under your name. It’s not something I do, in either direction. What I can do is read your draft and tell you what’s not working, in as much detail as you want. If it helps, I can go through the structure with you before you start.

   I’d rather not draft this for your signature. I’m happy to write the section that covers my own work and be named for it, or to read and comment on yours. If the deadline is the problem, tell me and I’ll help you work out what can be cut.

   Do not. Do it once as a favour. The second request always arrives, and it arrives with the first one as precedent.

9. 09 · You are being moved down the author order

   What is actually being asked

   For you to accept a smaller record of work you did.

   The position

   Ask for the reasoning against the contribution, in writing, without accusation.

   I’d like to talk about the author order before we submit. My understanding was that I’d be [POSITION], on the basis of [SPECIFIC CONTRIBUTIONS: the design, the data collection, the analysis, the first draft]. The current order has me at [POSITION]. Can you tell me what’s changed, so I understand the basis? If the contributions are being weighed differently, I’d like to see that written into the contribution statement either way.

   If the answer is unsatisfactory. Ask for the contribution statement to record precisely what each person did, and let the record be accurate even where the order is not. Then go to the integrity office if it matters to you, and it usually does.

   The thing nobody says out loud. Author order disputes are common, they are rarely about malice, and they are almost always about a conversation that never happened at the start. The scripts above are also the ones to send at the beginning of a project, when they cost nothing.

10. 10 · You are behind, and have to say so

    What is actually being asked

    Nothing. This one you initiate, which is why it feels worse and matters more.

    The position

    Lead with the new date. A supervisor’s fear is not that you are late, it is that you are hiding.

    I’m behind on [THING]. I’d told you [OLD DATE] and it isn’t going to happen. The realistic date is [NEW DATE]. What’s between here and there is [SPECIFIC REMAINING WORK]. [IF TRUE: The reason is [ONE SENTENCE].] I’d rather tell you now than send you something unfinished on [OLD DATE].

    Some of this is that I’ve not been at full capacity for the last few weeks. I’m dealing with it and I’ve spoken to [GP / STUDENT SERVICES / OCCUPATIONAL HEALTH]. I’m telling you because it affects the date, not because I need anything from you on it.

    You are not required to disclose a health reason. A date and a plan are a complete message, and “personal reasons” is a complete sentence.

11. 11 · A last-minute ask, arriving with urgency attached

    What is actually being asked

    For their planning failure to become your evening.

    The position

    Separate the urgency from the obligation. Somebody else’s deadline does not create yours.

    I can’t turn this round by [DATE]. I’ve got [ONE COMMITMENT] on the same days. If it’s still useful on [REALISTIC DATE], send it over and I’ll do it properly. If not, no hard feelings — come to me earlier next time and I’ll almost certainly say yes.

    The last clause is doing real work. It keeps the relationship, names the condition under which you are available, and teaches the pattern for next time. All three, in eleven words.

12. 12 · Out-of-hours messages

    What is actually being asked

    Usually nothing. Most out-of-hours messages are somebody clearing their own inbox, and the expectation of an immediate reply is one you supply yourself.

    The position

    State the pattern once, then follow it. A boundary announced and then broken is worse than none, because it teaches people to wait you out.

    I read messages during working hours and reply the next working day. If something is genuinely urgent, call me — that’s what the phone is for.

    I send email when it suits me and read it during working hours. Please don’t read the timestamp as an expectation of your time.

    I want to flag something small. I’ve been answering messages at [TIME] and I’m going to stop, so replies from me will come the next working day. Nothing behind it beyond wanting a workable week. If something needs a same-day answer, say so in the subject line and I’ll treat it that way.

13. 13 · The conversation at the start, that prevents half of the rest

    What is actually being asked

    Nothing yet, which is the point. This is the only script here you send when nothing is wrong.

    The position

    Put authorship, data and timelines in writing in week one, when nobody has anything to lose by agreeing.

    Before we start, can we write down three things, so neither of us has to reconstruct them in eighteen months? Authorship: who is on the paper, in what order, and what would change it. Data: who holds it, who can use it for what afterwards, and where it lives. Timeline: what we each owe by when, and what happens if one of us slips. Two paragraphs in an email is enough. I’ll draft it if that’s easier.

    We never wrote down the authorship and data arrangement, and I’d like to, now rather than at submission. My understanding is [YOUR UNDERSTANDING]. If that matches yours, reply and we have it on record. If it doesn’t, better to find out today.

    Why this is the highest-value page in the file. Scripts 02, 05 and 09 all exist because this conversation did not happen. Two paragraphs in week one removes most of the situations the rest of this file is written for.

14. 14 · A request to share unpublished data

    What is actually being asked

    For you to hand over the thing your next two papers depend on, usually framed as collegiality.

    The position

    Say yes to a defined use with a defined credit, or say not yet with a date. Never say a vague yes.

    Happy to talk about this. Before I send anything: what’s the specific analysis, what would the output be, and how would the contribution be recognised? I’m still working on [PAPERS], so [DATASET] isn’t available for general use until [DATE]. What I can do now is [SUBSET, OR SUMMARY STATISTICS, OR A COLLABORATION].

    Yes, with two things written down first: the specific use, and how it appears on the resulting paper. I ask because [DATASET] is the basis of my next two chapters, and once it’s in circulation I can’t un-circulate it. Once we have that in an email, I’ll send it the same day.

    If the data are not yours alone to give. Say that first. Consent terms, an ethics approval and a data-sharing agreement bind you regardless of who is asking, and “I’m not permitted to” is a complete answer with no negotiation attached.

15. 15 · You are accused of using AI, or pressured into using it

    What is actually being asked

    In the first case, for you to prove a negative. In the second, for you to carry a risk somebody else will not sign for.

    The position

    Answer with process rather than with indignation, and put any instruction to use a tool into writing.

    I want to answer this properly. [SECTION] is my own writing, and here is what I can show you: the version history in [FILE], the notes and reading I worked from, and I’m happy to talk through the argument of any paragraph you pick. On the detector result: I’d ask what its published false-positive rate is for writers working in a second language, because that is the group it misreads most. I’m not disputing the question, I’m disputing the evidence.

    Yes, and here is exactly where. I used [TOOL] for [SPECIFIC TASK — reference formatting, a first-pass literature search, a grammar check]. I did not use it to [DRAFT / ANALYSE / INTERPRET]. This is what my declaration says, and I’m happy to expand it.

    I’d rather not run [TASK] through [TOOL]. My concern is [DATA CONFIDENTIALITY / THE DECLARATION I’LL HAVE TO SIGN / THAT I CANNOT VERIFY THE OUTPUT]. If you’d still like me to, could you confirm it in an email, and tell me how you want it declared? I’ll do it on that basis.

    The mechanism in that last script. A request that survives being written down was fine. A request that evaporates when you ask for it in writing has told you what you needed to know, and neither outcome required an argument.

16. 16 · When your visa or your funding is attached to the answer

    What is actually being asked

    For you to accept a condition because the cost of refusing is your right to remain in the country.

    The position

    This is the one place in this file where the advice is not “say the sentence”. Get the facts from somebody whose job it is before you answer anything, because the real constraints are rarely the ones being implied.

    Thank you — I want to give you a proper answer rather than a fast one. My visa and my funding are tied to [PROGRAMME / SUPERVISION / SUBMISSION DATE], so I need to check what this changes before I agree to it. I’ll come back to you by [DATE, 3 TO 5 DAYS OUT].

    Then, in the same week, ask three people

    1. Your international student office or HR adviser. What are the actual conditions on your visa, in writing, and what does a change of supervisor, a suspension or a late submission do to them? The answer is often less catastrophic than the implication.
    2. Your funder’s handbook, or your funding officer. Interruption, extension and change-of-supervisor provisions exist in most schemes and are used more often than anybody says out loud.
    3. Your doctoral school or graduate school, not your supervisor. They handle these cases routinely and they are not the person the question is about.

    I’ve checked with [OFFICE] and my position is [FACT]. On that basis, what I can do is [SPECIFIC THING], and what I’m not able to do is [SPECIFIC THING]. I’d rather say that now than agree and fail to deliver.

    Why the sequence matters. A person who is told that their status depends on compliance rarely has that verified for them, and the implication does the work. Verified facts are the boundary here; the sentences come afterwards.

The reason is rarely the wording. Naming a new date is easy when you have a plan you believe in, and close to impossible when the schedule quietly stopped being real some weeks ago.

Dissertation Finisher — structure, weekly accountability, and a specialist who reads the work.

from €499/month· 6-month minimum on Lite and Full

It carries the Marathon Pause: once per commitment period, billing stops and the minimum term extends by the length of the pause. No penalty, and no explanation required. That last clause is there because the month you need to pause is the month you have least appetite for justifying yourself.

Or take the free one and nothing else. Momentum, the academic writing tracker, holds the streak and the pipeline board in your own browser. No account, no card, no paywalled tier. We would rather you used it and bought nothing than bought something you did not need.

No ghostwriting, ever · Human-led, PhD-signed