Should You Edit Old Journal Entries, or Annotate Them?
You reread something you wrote four years ago and the wince arrives before the memory does. The person in that entry is certain about a job that did not last, generous about someone who did not earn it, and spells "definitely" wrong twice in the same paragraph.
So your hand goes to the pen, or your cursor goes to the text, and the question shows up: should you edit old journal entries, or leave them exactly as they are?
There are really three questions inside that one, and they have different answers.
- Mechanics. A typo, a wrong date, a name spelled the way you first heard it said.
- Substance. The judgment, the tone, the conclusion you now think was wrong.
- Existence. Whether the entry belongs in the book at all.
Take them separately. The short answer is: fix the first freely, annotate the second, and delete the third rarely and on purpose.
What an old entry is actually good for
The events are recoverable elsewhere. Your calendar has the dates. Photos have the faces. What no other record holds is the state of your knowledge before the outcome, what you believed on a Tuesday while the Tuesday was still open.
That turns out to be the scarce thing, because memory quietly edits itself toward the ending.
Psychologist Baruch Fischhoff showed the shape of this in the mid-1970s. People who were told how an event turned out judged that outcome to have been more predictable than people who were not told. Knowing the ending makes the ending look inevitable. You do not experience this as distortion. You experience it as having known all along.
The written record is one of the few things that argues back. The morning after the Challenger explosion in 1986, Ulric Neisser and Nicole Harsch had students write down how they had heard the news, then asked the same students again years later. Many of the later accounts differed sharply from what those students had written themselves. Some, shown their own handwriting, trusted the memory over the page.
That is what an old journal entry is: the handwriting. Edit its substance and you have removed the only witness who was in the room before you knew how things turned out.
Mechanics: fix them, mostly
Not much is at stake in a typo. Correct it or leave it. Either convention works as long as you keep one.
Two cases deserve a pause.
- The error that carries information. You spelled a name wrong because you had only ever heard it out loud. That misspelling records how new the person was to you. Leave it and put the correction beside it.
- The date. Dates are load bearing in a journal. They are what makes sequence legible later. If you find a wrong one, add the correction rather than overwriting it, so the original mistake still shows and a future reader does not wonder whether a whole run of entries is offset.
Substance: annotate, do not overwrite
Revision is not a sin. Revising in place, with no trace, is the part to avoid, because it produces a document that presents itself as contemporaneous and is not.
Anne Frank is the clearest case. In 1944, after a radio broadcast asked people in the Netherlands to preserve their wartime diaries and letters, she began rewriting her own: reworking entries, cutting some, adding others, with publication in mind. Both her original and her revision survive, which is exactly why the revision costs nothing. You can read the diarist and the editor side by side, and neither one has to stand in for the other.
You can do the same thing at any scale. At the small end it is one dated line in a margin.
How an annotation layer works
- Date every annotation. "Sept 2026" in the margin is most of the trick. An undated note eventually becomes indistinguishable from the entry it sits beside.
- Keep it visually separate. A different pen, square brackets, an indent, or a plain "Later" prefix. You in a decade should be able to tell at a glance which layer is speaking.
- Add the fact, not the verdict. "He was already interviewing elsewhere" is worth having. "I was so naive" is not. It tells your future self nothing except that you were embarrassed on the day you happened to reread.
- Say what you did not know yet. The most useful annotation names the missing information: "I did not know the funding round had failed." That keeps the original entry reasonable given its inputs, which it usually was.
- Leave the grand summary alone. You will want to write a closing judgment on an entire era. Wait. Eras get re-judged. The annotation layer works because it stays small and cumulative.
On paper, use the margin, the back of the facing page, a sticky note, or an appendix at the back of the book keyed by date. In an app, where editing is frictionless and invisible, annotation is a discipline rather than a feature: append a dated block at the end of the entry instead of touching the body, and check that your export carries it out along with the text.
Yesterself is the journal that writes back. Every morning it sends you a letter from your own past writing, and your reply becomes tomorrow's. End-to-end encrypted, no account, no email — your archive is yours to export, free, forever.
Join the waiting listExistence: when deleting is the right call
Some entries should go, and pretending otherwise is its own kind of superstition. Reasonable grounds:
- It holds someone else's secret, told to you in confidence and written down without their knowledge.
- It describes a person in a way you would not defend, and it contains nothing else.
- You wrote it as a weapon, a rehearsal for something cruel, and rereading it only rehearses it again.
If you delete, treat it as a decision rather than a correction. Leave a line where it was: the date, and one sentence saying you removed it and roughly why. A gap you can explain is much better than a gap you cannot.
Two cautions. Journals are sometimes requested in legal proceedings, and destroying records once you know a dispute exists is a different situation with real consequences. This is not legal advice; if any part of that describes your circumstances, ask a lawyer before you delete anything.
And if rereading old entries reliably leaves you in a bad state, that is worth taking seriously rather than editing your way out of it. This is not medical advice, and a therapist or doctor is the right person to talk to. Closing the book and walking away is also allowed.
Where the software sits
Editing tends to be silent in software, and a journaling app is software. Rather than assume, check three things in whatever you use: whether it keeps any version history, whether an edit quietly changes the entry's original date, and whether your export carries annotations out along with the body text. The answers vary by app and they change between releases, so it is worth testing on one throwaway entry before you trust it with twenty years.
Yesterself is built around replying rather than revising. Its loop hands you a piece of your own past writing, and your response becomes a new entry of its own, so the old one keeps standing where it stood. The archive exports in open formats on every tier, the free one included, so the record stays yours to keep or to move somewhere else.
The one question worth asking
Before you change anything old, ask what you are changing it for.
If the answer is so I remember more accurately, annotate. Add the fact, dated, beside the original, and leave the original where it is.
If the answer is so I turn out to have been right, stop. The entry is already doing its job. It is showing you a person who did not know yet, which is the only condition any of us ever writes in.
Common questions
Is it wrong to fix typos in old journal entries?
No. Spelling and punctuation carry no evidence about who you were, so correct them or leave them as you prefer. The one exception is a mistake that is itself information, like a name you misspelled because you had only heard it spoken.
What is the difference between annotating and editing an entry?
Editing changes the original text so the old version disappears. Annotating adds a new, dated note beside the original and leaves the entry intact, which means a later reader can still see what you knew at the time and what you learned afterward.
Should I delete a journal entry I am ashamed of?
Shame alone is a weak reason, because the entry is usually showing you a person who did not have the information you have now. Deleting makes more sense when the entry holds someone else's confidence, or was written purely as a weapon. If you do delete, leave a dated line noting that you removed something.
How do I annotate an old entry in a journaling app?
Most apps have no annotation feature, so use a convention instead: append a block at the bottom of the entry that starts with today's date, and never change the body above it. Then run an export and confirm the appended note comes out with the text.
Should I rewrite my journal so it reads well for someone else later?
You can, but treat it as writing a second document rather than improving the first. Keep the original alongside the revision, the way both versions of Anne Frank's diary survive, so the record and the edited account can be read against each other.