The following recent question has attracted a fair amount of comments with anecdotal personal evidence about if and when the expression in question was originally used. There is even an answer based on such personal evidence.

I understand that statements such as the following may sound entertaining:

  • Yes, I'm sure that I heard that expression in the '70s, or in the '80s....sorry, my wife says it was in the '90s....so it must be the '90s.

  • I guess it was in Dallas where I first heard it used, or..wait..no, it was San Francisco, sorry.

but, if questions and answers are meant to be useful for present and future users, I think that considerations based on nothing else than personal memories should be avoided and possibly removed,

and while it is true that dictionaries tend to give dates about first usages which are later than actually spoken usages for obvious reasons, I think that the two quoted dictionaries are reliable and the dates that they suggest are to be taken seriously, unless written evidence that proves them wrong is produced.

  • 3
    I upvoted this question because I think it's a reasonable one to ask—but I also think that it is important to accord some weight to anecdotal evidence. For example, although I can't say with certainty when I first heard my Ontario-born grandmother use the term floster to refer to a small Danish pastry, I know that it was some time in the 1960s, and I know that I've never heard anyone else in Texas—or anywhere else in the United States where I've lived—use that term. So if anyone ever asks at EL&U about the origin and regional distribution of floster, I will cite my anecdotal evidence.
    – Sven Yargs
    Sep 23, 2018 at 23:08

6 Answers 6


I don't really see a problem here, and so I'm confused about your goal with this Meta question. It seems that you find these comments annoying, but you aren't obligated to respond to them, and I don't see how they do any real harm. In fact, I think that they contain some valuable information, although that's just my viewpoint.

Both citations and personal experience can be valid sources of information. The citations in your answer are not rendered worthless by the comments based on personal experience that suggest that the expression was in use earlier. The comments just add some additional information that readers are free to evaluate for themselves. You answer hasn't been downvoted by anyone, which to me seems to indicate that the users who left the comments don't think that your answer is useless.

Removing the comments would have to be done by a moderator (or by multiple flags being raised). If you want that to happen, I think your most effective course of action would be to raise a flag.

  • 1
    My point is that these comments, accurate or nor as they may be, don’t really add anything to the properly supported answer we are supposed to provide. They tend to be chatty and possibly confusing to external users. They are most likely based, even though in good faith, on false premises.
    – user 66974
    Sep 19, 2018 at 8:27
  • 10
    @user070221: I think users can decide for themselves how much stock to put in comments like this, so I'm not that worried about the possibility of them causing confusion.
    – herisson
    Sep 19, 2018 at 8:28
  • 1
    @it is a question of quality rather than quantity of desired information.
    – user 66974
    Sep 19, 2018 at 8:33
  • Users can remove comments if enough flags are raised. Moderators never even get to see them once the flags have worked: the system handles it all by itself.
    – Andrew Leach Mod
    Sep 20, 2018 at 6:19
  • 1
    @user070221 as one who often benefits from old questions and answers, I must respectfully disagree. In my opinion, anecdotal comments often help those who come later. Especially with some of the more technical stacks, anecdotal follow-up to an answer helps by providing more weight to your answer. For example, I'm more likely to trust your solution if it has already helped someone else.
    – Lumberjack
    Sep 21, 2018 at 18:37
  • 1
    Anecdotal comments (and even, in some cases, answers) are a bit like hearsay evidence, which - in my jurisdiction at least - is admissible in court but is assigned a lower value by the judge. In other words, it might not be as useful as "independently verified" evidence, but that doesn't mean it's not useful at all... Sep 23, 2018 at 6:24
  • @Chappo - I agree, but their “usefulness” is a bit dubious in a site that should provide solid evidence about English usage. Plus, how can users discriminate between reliable anecdotes and “fake” ones given that there is no way to verify them?
    – user 66974
    Sep 29, 2018 at 11:46
  • @user240918 Yes, the inability to verify is problematic, but is it really necessary in all cases? In a post today, for instance, I described how the street address 1/9 Park Lane is spoken in my city. It's useful information, but how do you verify it? What evidence would I cite? I think the hurdle for answers has to be higher than for comments, but we wouldn't want it so high that valuable detail is precluded. Sep 29, 2018 at 12:24

I could argue that comments like these are not against the rules because they're "suggestions for improvement", but I think that would be missing the point. (An answer only supported by anecdotal evidence such as this would be a different story, but I don't see that being a problem here, since the only anecdotal answer, the one that gives an "answer" of "I never heard this", is heavily downvoted.)

Comments like these shouldn't be deleted until they are made obsolete (e.g. if someone someone verifies it with evidence in an answer). I'm aware of all the relevant reasons we shouldn't keep comments like these (due to the way the Q&A system works or is supposed to work), but I think I'd rather keep them to help in the name of etymological research*. Plus, there's nothing particularly dangerous about these comments (unlike at a site such as Interpersonal Skills). Nor is it always easy enough to find an answer with sources to back up these comments. And how much stock is put in a comment like this that has "memory" as a source is up to each individual. In any case, it's very unlikely that we'd get any troll comments, since it requires 50 rep to do so.

* There's one compelling reason why I feel these comments help etymological research instead of hurting it: OED Appeals. Many Appeals are posted because anecdotal evidence suggests the date of first occurrence is earlier than what editors have been able to find. (For examples where they say this see here and here.)


The pieces of unsupported anecdotal evidence were not posted to criticise the OP's answer. They were testimonies from American English speakers who expressed incredulity about the dates supplied, not by the OP, but by the references cited by them in their answer.

Here are some other comments (some have since been deleted)

  • It's certainly decades older than this. I recall hearing it as a child, long before the 1990s.

  • @(X) I agree. I'm pretty sure I learned heard it growing up in the 60's and 70's

The OP says

I think that considerations based on nothing else than personal memories should be avoided and possibly removed

If competent native speakers and experienced contributors to the site cannot freely express their doubts about the validity of some posted references, in a perfectly civil manner, because their comments might get flagged then we really are taking CoC (code of Conduct) to its absurd limits.

  • Experienced contributors are more than welcome to express doubts about the validity of anything, but they should do that on a more solid base than just a juvenile memory.
    – user 66974
    Sep 20, 2018 at 16:34

The comment feature exists to help the community collaborate with the writer to improve the post by attaching:

  • clarifying questions
  • constructive suggestions and offers of relevant information
  • constructive explanations of actions taken – votes, edits, flags, etc. – when the actions might not be understood without an explanation

(Summarized from: https://english.stackexchange.com/help/privileges/comment)

The comments you describe are offers of relevant information. For now they should be left in place, to give the people involved a chance to reflect and decide what to do with the information. Later on, comments which will not be useful to future viewers can be deleted (or flagged for deletion).

  • 1
    Of any comments, these sort are exactly the kind that should be preserved. They are the few comments that are substantive: "here is some data that supports/falsifies the hypothesis". They don't belong as an answer (not yet possibly) but certainly should remain as supporting examples.
    – Mitch
    Sep 19, 2018 at 17:06
  • @Mitch That sort of thing is my third bullet. "Here's the reason why I upvoted (or downvoted)". Potentially very useful to future viewers.
    – MetaEd
    Sep 19, 2018 at 19:18
  • I suggest upticking them if they seem important and flag them for deletion if they are badly flawed.
    – MetaEd
    Sep 19, 2018 at 22:58
  • @JJJ I agree with my colleague. You were proposing to mechanically identify and mass-flag many types of unneeded comments, including "thank-yous". He rightly pointed out that while many comments may no longer be needed, mass flagging would interfere with moderation. That would make the site worse, not better.
    – MetaEd
    Sep 20, 2018 at 15:08
  • @JJJ Also bear in mind that mods do not normally make a determination of what information is right or wrong on the site. Mods will decline flags for wrong information. Normally flags are for circumstances where things have gone off the rails and require intervention to fix the problem. A comment containing a fact that you cannot verify does not need to be flagged.
    – MetaEd
    Sep 20, 2018 at 18:25

While anecdotal evidence is the least reliable, it's not worthless, especially when it comes to language and history. Where do you think some of the stuff that gets written down comes from if not from people recalling and sharing things they've experienced? We all know slang was in use long before a lexicographer thought to create an entry for it, although now that so many of us have access to the Internet, we should be able to do more accurate tracking of the origin and spread of slang.

Sure, the people posting those comments could have a faulty or vague memory, but as long as it's obvious it's anecdotal and less reliable, it's not harmful. What's wrong with knowing that there may be some significant differences in the first date that we can validate a usage and when it was actually being used in spoken English? We might not be able to say "it was in use in 1968", but we can say it may have been in use at least 20 years before it got recorded in the dictionary.

When I was in school, one of our history projects was to interview our family members and record their recollections about our family's history, then see what we could corroborate with other sources. Hard data is important, but anecdotes can help us discover facts so they can be verified, and help us put them into context, which helps us remember them. Anecdotes may not be suitable support for answers, but as comments I think they can be worthwhile additions.

  • 1
    While I agree that telling stories and experience does no real harm, as a language site which aspires at providing reliable answers, indulging in personal anedctots doesn’t make this place more qualified but just more chatty. Anyway I do appreciate that comments have been moved to chat.
    – user 66974
    Sep 20, 2018 at 16:27

Anecdotal evidence like in your question should not be posted in comments. Some reasons:

  • Cannot be independently verified by other users or visitors.
  • Users cannot downvote the comments to express disagreement or other issues. Therefore, bad anecdotes remain on the site and are prone to mislead visitors.
  • Anecdotes lower the site's quality. For ELU to be successful it needs reliable information. If visitors need to double check everything it's no better than any other website out there.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .