How do I get back into a research project after months away?

Re-entry is hard for a reason that has nothing to do with forgetting facts. What disappears is the working context: the reasoning you were in the middle of, the option you had rejected last week, the sense of which loose end mattered. The files are all still there, which makes the difficulty feel irrational, and it is not.

Writing from memory rebuilds that context faster than reading does. Give yourself an hour and produce a page covering what the project claims, why it matters, what you had done, and what came next. Do not check anything while writing. Mark uncertainties with a question mark and keep going.

The page will be roughly seventy percent correct, and the marked gaps are your reading list. That is the whole trick: rather than rereading everything on the chance that something is missing, you find out precisely what is missing in an hour.

Then pick the smallest possible restart task. Not "finish chapter four" but "rewrite the first paragraph of chapter four". Re-entry after a long break fails more often from an oversized first task than from lost knowledge, and one completed small thing restores more momentum than a day of preparation.

If the project has an analysis component, make the smallest task a rerun rather than a read. The reproducibility standard Sandve and colleagues set is that you record programs, parameters, and manual steps well enough to approximately reproduce a result "in a year or so" (Sandve et al., 2013, p. 2). Months away is exactly that test. Rerun one analysis, check which data version and settings it used, and write down what happened, including the attempts that failed. Their argument for this is time, not virtue: methods and results recorded this way get reused on new data instead of rebuilt (Sandve et al., 2013, p. 1).

Give that page a fixed place to live. Schnell's notebook rules are chronological entries, each one dated and labeled by subject, with a table of contents recording date, subject, and location, using labels as plain as "Research Review with Dr. Williams" or "Thoughts on Project X" (Schnell, 2015, p. 2). Make one entry called "Project restart" holding the last completed task, the open questions, the files that matter, and the next decision. In an electronic notebook the contents list builds itself (Schnell, 2015, p. 4). Before you reorganize anything, check the rules you work under, because research records are often governed by institutional procedure and notebooks produced with institutional funding belong to the institution (Schnell, 2015, p. 1).

Do I have to reread everything?

No. Reread the three or four works your current argument depends on and nothing else until a specific gap sends you back. Rereading the whole library is the most common re-entry strategy and the slowest, and it postpones the writing that would show you what you still know.

The instinct behind it is understandable. Rereading feels safe, it is unambiguously productive, and it does not risk producing bad writing. That is exactly why it becomes a way of avoiding the return rather than accomplishing it.

Trust your notes to be the record. If they are not good enough to support this, that is a finding about the note system worth acting on before the next break rather than a reason to reread now.

What should I write down before a planned break?

A handover note to yourself: what the argument currently is, what you were about to do next and why, what you had ruled out, and which three files matter. Twenty minutes before you stop saves several days on return, and nobody regrets having written it.

The rejected options are the part people leave out and miss most. Without them you will rediscover a dead end, spend a week on it, and reach the same conclusion you already reached.

Write it addressed to a colleague rather than to yourself. Writing for an audience forces you to state things you would otherwise assume, and after four months you are close enough to a stranger for that to be the right register.

If the work depends on a supervisor, a lab, a data steward, or a shared system, the handover note has a second reader. Vandendorpe and colleagues, writing about getting electronic lab notebooks adopted, recommend staged monitoring, hands-on training or direct visits where they are needed, systematic feedback through questionnaires, interviews, or group discussions, and open communication, because continuing support matters more than the initial setup (Vandendorpe et al., 2024, pp. 5-6). They also list templates for protocols and workflows, metadata and tags, local operating procedures, and community resources as the things that keep records usable (Vandendorpe et al., 2024, p. 6). Turn your restart page into a one-page agenda for a meeting: what you recovered, what is still ambiguous, what access or training you need, and what your supervisor has to decide. None of these three sources studied students returning after an absence, and they emphasize different things, Schnell record structure, Sandve reproducible analysis, Vandendorpe institutional support. Treat them as a checklist for the return, not as proof that any single routine restores productivity.