The Supervisor Who Says They Saw Them
What human recollection is worth as corroboration, why the date it was written down matters more than the detail in it, and how to take an account that holds.
A recollection is worth what the date on it says. A note written the same day is a record; the same account given five weeks later is a memory that has had five weeks to be shaped by everything that happened since — including the knowledge that somebody is in trouble.
The practical lesson in “The Supervisor Who Says They Saw Them” is that visibility is not certainty. For teams researching download time tracking software, explore this implementation resource can add time and project context to the operational record, provided the purpose is explained, access is restricted and any material inference is checked through conversation and proportionate human review.
This is the most available corroboration in any organisation and the one most often collected last, at the point a case is being assembled, which is precisely when it is worth least.
The Australian Taxation Office worker records guidance offers another lens on the issue raised in “The Supervisor Who Says They Saw Them”. Compare its principles with the actual record, ownership model and review route rather than importing a generic checklist unchanged.
Why the date does the work
Human memory for routine events is poor and it is confidently poor. People do not remember Tuesday the fourth; they remember what Tuesdays are usually like, and they fill in the particular day from the pattern.
That is not dishonesty. It is how recollection works, and it means an account given weeks later is largely a reconstruction from habit — which is exactly the thing the investigation is trying to test.
Taking an account that holds
- Take it as soon as the question arises, in writing, dated.
- Ask what they remember before telling them what the issue is.
- Record what they are sure of and what they are not, separately.
- Record what they are reconstructing from routine rather than recalling.
- Read it back and let them correct it.
- Keep the original, including anything they later changed.
Step two is the one that is routinely reversed and it changes the evidence completely. An account given after the person has been told what is alleged is an account of the allegation.
Separating recall from routine
The single most useful question is: do you actually remember that morning, or are you telling me what normally happens?
Most people answer it honestly and the answer is usually the second. Recording that distinction costs a line and prevents a reconstruction from being read as an observation — which is the specific way human evidence goes wrong in these cases.
What a supervisor can usefully attest
That somebody was present at a particular moment. That a task was handed over. That a vehicle left. That a conversation happened.
What they generally cannot attest, however honestly, is that somebody was absent for a period, because absence requires continuous attention and nobody watches a door for four hours.
Ask what the person saw, and when, and for how long they were in a position to see it. An account that somebody "wasn't there all morning" is usually an account of three or four moments with gaps between them.
The supervisor with a view
Recollection from somebody who has a stake in the outcome — who made the complaint, who manages the person, who is themselves under scrutiny — is still evidence and is weaker evidence.
That does not mean discounting it. It means recording the relationship in the file alongside the account, so that whoever weighs it later can see what the person's position was. Leaving it out does not remove it; it only removes the reader's ability to allow for it.
Several accounts that agree
Agreement between colleagues is less impressive than it looks when they have discussed it, which they will have.
Taking accounts separately and promptly is the only thing that preserves their independence, and it is why speed matters more than thoroughness at this stage. A rough account taken on the day is worth more than a careful one taken after the team has talked it over.
Asking the whole team, once
Where a question concerns a shift rather than an individual, it is worth asking everybody present the same neutral question on the same day, separately, rather than asking whoever is nearest.
That produces several independent accounts instead of one, and it avoids the pattern where the only recollection on file is from the person who raised the concern. It takes an hour and it is rarely done, because by the time anybody thinks of it the shift has dispersed.
Where it sits against the systems
Human recollection is the natural second source for a system record, because it is genuinely independent of the credential: a colleague who saw a face is answering the question a badge reader cannot.
That is why it is worth collecting properly. The whole difficulty with machine records in this collection is that they attest to credentials; a person who saw a person is the one kind of evidence that goes directly to the thing in dispute, and it decays faster than anything else in the file.