The value of field notes as evidence comes almost entirely from whenthey were made. A record created at the instrument, at the moment of observation, carries weight precisely because it wasn't assembled later with the benefit of knowing how things turned out. That's the whole basis of the contemporaneous record — and it's fragile in a specific way.
A set of notes showing clean, deliberate corrections is a normal working document. A set of notes showing an erasure, a scribbled-out block, or a number written over another number invites a different question entirely: what else was changed, and when? Once that question is on the table, it attaches to the whole book, not just the one entry.
The Convention, and Why Each Part of It Exists
The standard practice for correcting paper field notes is narrow and deliberate:
- Single line through the error — struck through, not obliterated. The original must remain legible.
- Write the correction adjacent — above, beside, or on the next line, not on top of what was there.
- Initial and date it — who made the correction and when.
- Never erase, white-out, scribble over, or overwrite.
- Never remove a page or renumber pages. A missing page is worse than any error that was on it.
The “keep the original legible” rule is the one people find fussy, and it's the one that matters most. The point isn't to preserve a record of your mistakes for their own sake — it's that the original value may turn out to be meaningful later. A rod reading struck through and corrected from 4.82 to 4.62 tells a reader that the reading was taken, questioned, and revised. The same cell scrubbed clean and rewritten as 4.62 tells them nothing happened at all, which is a claim the notes can't actually support.
Corrections Made Later, Away From the Site
There's a meaningful line between two things that look similar on the page.
Correcting a transcription error is defensible.You wrote the instrument height in the wrong column, or transposed digits copying a published benchmark elevation from one page to another. The underlying observation is intact; the writing-down of it was wrong. Strike, correct, initial, date, and — if it's a later date than the original entry — say so.
Changing an observation from memory is not.You cannot recall a measurement more accurately than you recorded it. If a value looks wrong back at the office and there's nothing to verify it against, the honest move is a note documenting the discrepancy — not a quiet replacement with what the number probably should have been. A field note that says “BS 4.82 does not reconcile with closure; re-observation recommended” is a professional record. The same book with 4.82 silently changed to a value that closes is something else.
This is the same failure mode as filling in notes after the fact — the record stops describing what happened and starts describing what would have been convenient.
The Digital Problem Nobody Brings Up
Here's the uncomfortable part, and it applies to this entire software category rather than to any one product: a digital field note that can be edited silently is weaker evidence than paper, not stronger.
Paper has a useful property — it physically resists concealment. An erasure leaves a mark. A rewritten page looks rewritten. Ink applied months later often looks different from ink applied at the instrument. None of that is foolproof, but it means alteration tends to leave traces.
A database row has no such property. A rod reading entered in May can be changed in November with no visible evidence that it ever held a different value, and the record will look exactly as pristine as one that was never touched. That is a genuine regression in evidentiary quality, and it's worth being clear-eyed about rather than pretending digital is strictly an upgrade — a point worth weighing alongside the real advantages covered in our paper-versus-digital comparison.
What closes the gap is an audit trail. When you're evaluating any digital field notes system — ours included — these are the questions that actually matter for this problem:
- Are entries timestamped at the moment of capture, or only when the record is saved or synced?
- If a value is edited after the fact, is the prior value retained and visible, or overwritten?
- Is there a record of who made a change and when?
- Can an entry be deleted outright, and does anything remain to show it existed?
A system that keeps prior values and attributes changes reproduces what the strike-through convention does on paper. One that doesn't is asking you to take the current state of the record on faith.
What a Good Correction Actually Looks Like
On paper, in the moment:
FS 4.82 4.82 4.62 — misread, re-observed immediately. CG 9/28
Same day, same crew, original legible, reason stated, initialed and dated. Anyone reading that book in five years knows exactly what happened and has no reason to wonder about the rest of it.
A correction made two weeks later, in the office, gets the same treatment plus one more element — an explicit acknowledgment of the gap: “Corrected 10/12 from field sheet; original entry transposed.”Nobody is fooled by a correction that pretends to be contemporaneous, and nobody objects to one that doesn't.
Quick Reference
- Single strike-through, original legible — never erase, white-out, or overwrite
- Correction written adjacent, with initials and date
- Never remove or renumber pages
- Transcription errors can be corrected; observations cannot be corrected from memory
- If a value can't be verified, document the discrepancy rather than replacing it
- Corrections made later should say so — the date gap is not a problem, concealing it is
- For digital notes, an audit trail that retains prior values is what replaces the physical evidence paper provides