<whoosh>
Do you have an agenda, too? Why is “the data is” an error?
I don’t say “datum” ever, but I will write things like “the data are”.
I’m actually a reformed singlar-data user. With training in mechanical engineering, I heard and used datum frequently (a common term for reference points and planes in CAD-like software). Now, working around geologists, I hear and read data are all the time, and switched to that camp not too long ago.
I use data, media, and criteria as plurals. I haven’t had much of a chance to use datum.
I wouldn’t say I have an “agenda.” Sal asked if anyone is still holding out.
My Webster’ New World Dictionary says n.pl [often construed as sing.]
My American Heritage Dictionary says pl.n. Data is now used both as a plural and a singular collective.
And, as noted earlier, my AP Stylebook says it takes a plural verb, except when used as a collective noun.
When I type “data plural” into Google I get more than 8,000,000 pages, so I 'd say there’s a fair amount of discussion about it.
What do you mean, “anyone holding out”? There’s no holding out involved. Data, media, and criteria are plural. Datum, medium, and criterion are singular. The sentence “my data is bad” is just plain wrong. Yeah, maybe it’s pedantic, but I never adopted the “everybody else is doing it wrong, so I will, too” philosophy.
And, being pedantic, I’d say “The shoal [of fish] is swimming.”
Whoosh, whoosh, whoosh, whoosh, whoosh.
Why didn’t you write “I wouldn’t say I have an agendum”? After all, that’s the Latin singular for “agenda.”
Answer: you’re not talking Latin. You’re talking English. The fact that “data” is plural in Latin is irrelevant.
Data/Datum
Agenda/Agendum
What’s wrong with keeping the plural/singular bits straight?
I don’t usually refer to datum singularly, so I don’t use the word very often. But it’s nice when someone is aware of the distinction.
I generally say data are, rather than data is. When I must refer to the entire set of data, I would say the data are because the entire set is still a group.
When I have to refer to more than one medium, I call them psychics. Then I make fun of them.
Tris
I still use data as a plural word. Its partly because the style manual for the professional journal in my field demands that we use data as a plural, so its been drummed into me. The phrases “the data is” or “the data shows” just looks plain wrong to me now.
How 'bout this:
You keep fish in an aquarium.
But who ever says “I have two aquaria” ?
I studied Latin in my schoolboy days (and actually liked it!). But nowadays, I speak English. Plurals end in s, not a.
“the data is accurate” , “the news media is biased” etc.
It would have sounded grammatically incorrect to ancient Romans.
But they’re all dead, and I’m not. 
Likewise, Persian borrowed tons of words from Arabic, including the complex “broken plurals,” which to a Persian who doesn’t know Arabic, are not recognizable as plurals. Persian, like English, forms plurals by adding suffixes. Arabic usually makes plurals (called “broken” because the internal structure of the word is broken up and the pieces rearranged) by changing the pattern of vowels while keeping the three consonants of the root.
So just like data, media, agenda, and (still illegitimately)criteria, these Arabic plurals became treated in Persian grammar as singulars. Example: Arabic ‘ulamâ’ means scholars, the plural of the singular form ‘âlim. Because the ‘ulamâ’ held a role in Islamic society as a discinct class of people, the plural form was commonly used even in non-Arabic speaking countries. Apparently, as the more familiar form of the word, it was adopted for the general meaning, so its plurality became unmarked. Then in Persian people would say sentences like Az yak ‘olamâ be-porsid agar dânestan khvâhi (Ask an ‘ulamâ’ if you want to know. Yak means ‘one’ or ‘a’).
Which is the equivalent of saying in English “I just need one criteria.” But over time, everyone got so used to talking that way, it became acceptable Persian grammar. Later, many languages all across Asia and elsewhere, like Turkish, Urdu, Uzbek, etc. became heavily influenced by Persian, and they took up this use of the plural form with singular meaning.
Which reminds me, even in English I sometimes hear “a mujahidin” – this is an Arabic “sound” (i.e. not broken) plural, made by simply adding a suffix. It’s like the Hebrew -im in cherubim.
People are still insisting that “data” is a plural noun? Those are news!
I tend to think of “data” as uncountable but “criteria” as countable, so I use “criterion” where relevant. On the other hand, I’m a fussy sod who will write “12m” for midday, even if I end up having to explain myself. 
Yep. From the Monthly Notices of the Royal Astronomical Society style manual:
Maybe its a Brit thing?
Well, the British do say stuff like “the government are…” They all just need to learn us how to conjugate some.
Here’s something confusing (from Kunilou’s post above): “My AP Stylebook says it takes a plural verb, except when used as a collective noun.” I’m having trouble making the distinction. (Though, to revert to my previous example, I think there are times when British people say, “The government is…” if they’re thinking of the government as a collective entity, as opposed to an aggregation of individuals – again, a distinction I have trouble making.)
Don’t think so. I think it’s really more of a stodgy-crumudgeon vs. non-stodgy crumodgeon thing.
A random and unscientific sampling of American institutions that prefer data to be used as a plural noun:
American Psychological Association
University of Rocheter
University of Montana
University of Texas System
Medical Library Association
And, just for you, Angua, AJ & ApJ. Icarus has no notice of it in their guidelines for authors, but the current editor prefers the plural (P. D. Nicholson, personal communication, 2005). But, then, he’s Australian . . .
As a scientist, I just reflexively used data as a plural. Early on, I had the singular usage corrected often enough by reviewers or editors that it’s natural now. If I used it as a singular in a professional context, I would be ridiculed as ignorant.
“Data” is plural in English too, according to most (as far as I know, all) dictionaries, as well as both Chicago and AP stylebooks (even if some may also allow its usage in the singular in some constructions).
Merriam-Webster has this to say:
While I would certainly not say that “data is” is an error, the idea that anyone is “holding out” for “data are” is rather bizarre, since this is the standard construction in many contexts, and certainly the most widely accepted one by dictionaries and stylebooks.
“Data are” and “data is” are both correct in English. However, using “data is” will make you appear ignorant to many people. Using “data are” will make you appear stuffy to other people. The choice is yours! 
I never understood the attitude being shown by some of the posters in this thread. It comes across as “A lot of people use this word wrong (e.g., data as singular), so I’m going to use it wrong, too. If you use it right, you’re a stuffy old fool.”
Fine. If you want to misspeak because all of your friends misspeak, then by all means do so. But why ridicule or insult the people who still say it correctly?
(raises hand tentatively) That’s the way I learned it. One aquarium. Two aquaria. I’ve never heard an educated adult say “aquariums” before. Is it common?
Invisible Wombat, as a historical linguist, I can say that as a prescriptivist you are fighting a losing battle. Living language does not behave as prescriptivists say it ought. It behaves however it behaves and that’s that. The only cases in which prescriptivists actually get their wish, and have all the rules locked down, are dead languages. Would you want to kill English? Murderer! 