Wiz, I don't disagree with Quinion's first quote above and would say that his comments on ‘country of origin and dating’ for slang being based on first recorded use’ is not all that shocking to anyone who follows word and phrase origins. Dating (along with a country) is often given as a range surrounding that first known appearance (see Random House Unabridged Dictionary and Cassell’s Dictionary of Slang), which is often cribbed from the OED, and seems to me to be exactly what it says it is and nothing more – first known/recorded appearance in print (sometimes accompanied by country). And I’m not aware of any reliable sources ever saying it was anything but that. The folks who are ‘mislead’ are those who don’t know what those ‘first appearance’ dates mean – and it is unfortunate that they don’t and that they get a wrong impression. And I do agree that it is often the case that it is not clearly defined by the authors – not that most readers look at the explanatory notes they provide at the beginning of a dictionary anyway.
I think that the reason that dictionaries, both slang and standard, often resort to this date is that guesses as to when and where a word or phrase was ‘first spoken’ are mostly too flimsy to hang one’s hat on and too unwieldy to adopt as a guiding principle for an entire dictionary. The labor and difficulty, and unreliability involved in such guesses, especially when dealing with thousands, tens of thousands (e.g. Cassell’s Dictionary of Slang, Barnhart Concise Dictionary of Etymology), or even hundreds of thousands of entries (e.g. OED) make such a choice undesirable and impractical.
So what’s a lexicographer to do? First appearance in print seems like a reasonable and workable dating system (for both slang and Standard English). And I suppose I would rather have a date that was at least well-defined, although perhaps somewhat conservative due to the time lapse between first spoken and first printed, than have no date at all, or one based on someone’s wild-assed guesses of the mostly impossible-to-determine ‘date first spoken.’
I was very disappointed when Merriam-Webster Online pulled the dates from all of their entries a few years ago. Their philosophy on that one might have been that, besides saving a tremendous amount of research effort, time, and bucks, they would rather have no dates at all than be caught in what appeared to be a mistake when an earlier date was found. The OED has no such qualms and is updating its electronic version to account for new information (often provided by readers) all the time. I, for example, regularly find appearances in print that predate that of the OED (and submit them), but I don’t get overly upset and complain that I was being mislead. I consider the entries for what they are – best attempts, thus far, at finding the first appearance in print, and at the ready at any moment to be replaced by any newer discoveries.
Ken – April 2, 2006