Please do not re-nominate for verification without comprehensive reasons for doing so.
Paul G tagged this entry for clean up due to the fact that, although it had nine citations, all of them were for its non-standard uses as a singular or an uncountable noun. I’ve since sorted by sense and moved the citations to a subpage. Obviously, I’m not disputing that criteria exists as the plural of criterion — all I’m saying is that it could do with some citations to prove such a usage. I don’t have the time at the moment to add the citations myself, so I invite everyone else to add a few. There should really be at least ten for the correct usage. † Raifʻhār Doremítzwr 15:23, 20 June 2007 (UTC)
Any thoughts on criterior? I've marked it as a misspelling or nonstandard, but there are so many google books hits it might be considered a legit alternative. Kappa 00:29, 21 June 2007 (UTC)
Most of those are scannos, but a number are genuine somethings … typos for criterion? misspellings/hypercorrections of criteria by speakers of non-rhotic dialects? The surprising thing is that criterier gets very few hits; there seems to be some consensus that even if you spell it with an R, you need an O. :-P Regardless, I definitely think it's either nonstandard or a misspelling; it gets a negligible number of hits compared to criterion or criteria, and I imagine that people who write criterior do so for lack of awareness, not because they consider it an equal alternative. (It's hard to be sure, though.) —RuakhTALK 04:19, 21 June 2007 (UTC)
Yes, the only sense I could make of it was that it might be a misspelling by speakers of non-rhotic dialects, as Ruakh suggests. If it sees consistent use (as opposed to being used once in a book that elsewhere uses criteria), mark it as a misspelling. — Beobach972 15:59, 21 June 2007 (UTC)
(Or a scanno/typo for criterion. — Beobach972 16:11, 21 June 2007 (UTC))
I have placed quotations in the main entry and in the citations subpage. — Beobach972 18:39, 22 June 2007 (UTC)
To be a descriptive dictionary, we should therefore indicate that criteria is the accepted singular, while criterion is a product of etymological hyper-correction? --Connel MacKenzie 21:39, 21 June 2007 (UTC)
Not at all. :-) Descriptivism and prescriptivism are in agreement here. If you search b.g.c. for "criteria are" and for "criteria is", you'll see that not only does the former get more than three times the hits the latter does, but also while most of the hits for the former are indeed using criteria as the subject of are, most of the latter are not actually using criteria as the subject of is. —RuakhTALK 19:13, 22 June 2007 (UTC)
What exactly is the threshold at which a misspelling is deemed “common” and thereby qualifies for an entry? † Raifʻhār Doremítzwr 14:59, 22 June 2007 (UTC)
We haven't developed a strict threshold, that I know of. Dmh's suggestion was >1% as common, but <10% as common. Both "relative tests" were meant as measures of web-search frequency (which has its own inherent problems, but works as a starting point for discussion.) While that works in some cases, in others it is obviously inappropriate. Do you have any suggestions for narrowing down what we mean into formal criteria? To date, I haven't seen a pattern emerge (from empirical observations of RFD/RFV/RFC.) To take a horrible example, http://www.google.com/search?q=colour vs. http://www.google.com/search?q=color implies that 'colour' is more than a misspelling, but instead is an alternative spelling. (Note: this is an example of why the numeric approach does not always work, especially as an absolute measure. Such numbers cannot be taken as absolutes, rather, as indicators, in context of what else is known about the terms in question.)
The 1% criterion doesn't always work though. And certainly any effort to generate combinations that simply meet that criterion will (again) be met with tremendous resistance. Proposing something more concrete with two dozen examples (or so) might generate enough discussion to eventually lead to an actual vote. --Connel MacKenzie 17:26, 22 June 2007 (UTC)
Whatever our statistical tests and criteria, the important thing — if we are to be descriptive — is that the words we call misspellings are indeed misspellings and the words we call alternative spellings are indeed alternative spellings. That is, it is important that our entries reflect writers’ intentions. IMHO, we are far more likely to achieve such accurate representation of usage by using our rational faculties to make educated guesses of writers’ intentions whilst running the risk of introducing subjective bias than we are by applying a small handful of restrictive and simplistic numerical tests in an effort to be completely objective and NPOV.
Of course, “rational faculties” are generally pretty objective. The kinds of thing I was thinking of were likelihood of typographical error (as I explained in the usuress discussion, but which rationale would ideally replaced by a peer-reviewed academic paper on the subject), failure to follow an unusual rule (as in the cases of geneology and hypernym), unintentional hypercorrection (e.g.: whom used nominatively and it’s as a possessive), and so on. In very little time, I’m sure that the En.Wiktionary community could formulate an effective set of principles to deciding what is and what isn’t a misspelling. † Raifʻhār Doremítzwr 06:30, 7 July 2007 (UTC)
Alright, let's explicitly distinguish alternative spellings from misspellings : an alternative spelling for a word is a spelling which is consistently used and has enough usage to meet CFI. A misspelling is a word that is not consistently used : if a book spells the wordd-o-g-g once, but otherwise spells it d-o-g, we can assume that d-o-g-g is a misspelling. That's my understanding. Now, here's my suggestion :
For a given misspelling to be included, I propose that it must meet one of the following : (a) it is used (in the inconsistent manner described above) in a well known work. My second test, (b), is labour-intensive : the correct spelling of the word has at least 90 examples of usage (we don't want misspellings of rare words), and the misspelling at least 30. That is too labour-intensive to be of much use, or garner much support, so I know it needs to be revised...
Also, I suggest that if a word merits inclusion as an alternative spelling of X, we'll never include a separate definition line that labels it a misspelling of X, even if it meets the aforementioned tests. — Beobach972 18:09, 22 June 2007 (UTC)
That sounds like a good way to distinguish typos, but I don't see how it can distinguish misspellings? —RuakhTALK 18:22, 22 June 2007 (UTC)
You're right, you know; I didn't think about it like that... that may only work for typographical errors. — Beobach972 18:43, 22 June 2007 (UTC)
Why would you want to exclude the indication of likely misspellings? And why the sudden jump to 33.3% (which would pretty clearly imply it is an alternate spelling, not a misspelling, anyhow.) We might have a better chance of describing a statistical measure of misspellings, if we had a clearer definition of what we mean by "misspelling" vs. "typo" vs. "mis-construction." Do we even have examples of each, that we can use for comparison? --Connel MacKenzie 16:24, 7 July 2007 (UTC)
What is the etymological relevance of criteria vs criterions? I guess this isn't isolated to these words, but there are no purple elephants. —This unsigned comment was added by 188.8.131.52 (talk) at 15:44, 13 October 2009 (UTC).
What do you mean by “etymological relevance”? The difference between criteria and criterions is that, whilst the former derives from the Ancient Greekκριτήρια (kritḗria) (the nominative plural form of κριτήριον (kritḗrion, “criterion”)), the latter has been Anglicised as criterion + -s (plural suffix). Traditionally, many Ancient Greek words, including those (like criterion) ending in -on, were either Latinised (e.g., the Græcian -on became the Latinate -um) or fully Anglicised by discarding the case ending; in the case of criterion, both these things happened, creating the rare forms criterium and criterie, critery. Philologically speaking, the proper plural of criterion and criterium is criteria, whereas the proper plural of criterie and critery is criteries. I don’t get what you mean by “purple elephants”. Does all that answer your question? †﴾(u):Raifʻhār(t):Doremítzwr﴿ 17:44, 13 October 2009 (UTC)