Grammar Snobbery

[rochvelleth]

It is generally known that I am the worst kind of linguistic pedant. I make no apologies for this, most of the time.

Anyway, two issues recently came up for discussion. Please help me discuss.

'Data' is plural

I would generally insist that 'data' should be treated as plural in all cases (though I approve of vyvyan's treatment of it as a mass noun, it'

Comments

[the_lady_lily]

You know, I've never felt too strongly about the split infinitive, mainly because I don't think I've ever been formally told what one is. It is just this bogey grammar monster that sits in a closet waiting to jump out at me in undergraduate essays, I think. I tend to be led far more by whether the construction of the sentence is intuitive and easy to understand, which may mean I pick out split infinitives as clumsy and unnecessarily confusing rather than as split infinitives per se.

[rochvelleth]

I tend to be led far more by whether the construction of the sentence is intuitive and easy to understand

Ah, yes, I definitely agree with that sort of approach. Actually, I wilfully break grammar 'rules' like this just for effect in fiction, or even in unimportant academic work.

[cartesiandaemon]

1. I like the mass noun theory. I know *sometimes* data is being the plural of datum, but I don't think that's how I think of it most often. Is there a word you'd rather I use when I mean that?

For instance, if you were comparing data, you would probably want to think about the amount of information rather than the number of datums. If A categorieses his datums in pairs, and B doesn't, I would think of B as having twice as many datums, but the same amount of data.

I use fora sometimes because it's more fun, but unless it's really established, I have to admit using 's' plurals are more communicative.

2. I agree you should not split infinitives if you're going to be penalised for it. But I don't think argument from authority is necessarily good.

Does splitting an infinitve sound worse? Not imho, but let's argue that in a moment. I think it's often cleaner and tidier.

To the point, I don't think it can *always* sound better, and that it is a point of style, so shouldn't be considered a grammatical rule.

[rochvelleth]

Well, that's why I think the mass noun thing works really - it's difficult to argue with what you've said. But, I mean... well data *is* used as a standard plural though. Isn't it? Is it really only classicists who use it like that?

[vyvyan]

I have a mini-corpus of English academic published texts on CD (from the British National Corpus), which I've just done a quick search of, looking for uses of "data". There were more hits than I had time to look through, so I checked every 5th hit, producing 159 instances. Only 25 of these were clearly either singular (mass) or plural (count) instances; 13 were plural and 12 were singular. Some examples of singular uses (in recent accepted academic published texts):

Positivist research has generated much [[data]] about specific relationships between individual or social characteristics...
... discuss, not 'rules' but regularities, simply because his [[data]] constantly exemplifies non-categorial phenomena.
Equation (10.10) can be used to analyse [[data]] such as that in figure 8.7.
The [[data]] corresponds precisely to the situation described by Ashton et al ...

While it's clearly not just classicists who use plural "data", I think this gives evidence against the claim that Academics in particular will pick up on it if you don't

[cartesiandaemon]

Also, if I *did* mean "plural of datum" I think I *would* use data, but I don't think that is what I mean most often, if you see what I mean.

These are just shibboleths, aren't they?

[ewx]

I don't buy the etymological arguments concerning data - there are lots of other borrowings into English which have been mangled one way or another compared to whatever language they came from. Nonethless "a data" would certainly sound odd to my ears.

"To go boldly" sounds far more clunky to me than "to boldly go", not just in that particular famous case but also often in the general case where an adverb is shifted outside the infinite. I note in passing that I've never heard anyone complaining about "split negatives" in French...

A big point in favour of enforcing the no split infinitives rule is that academics will pick up on it

"A code" for a "a program" sounds quite bizarre to my ears (trained in this instance by many years of hobby and professional software development), despite apparently being quite a popular usage in some academic circles. Which, with the other remarks about academics picking up on particular usages, I think emphasizes that academics are just another language community (or communities), with their own

Re: These are just shibboleths, aren't they?

[ewx]

Oh, and "boldly to go" sounds even worse.

Re: These are just shibboleths, aren't they?

[rochvelleth]

Well, I *did* mention I was strange when I said I liked it :) It probably is Ciceronian influence that makes me like that sort of word order actually...

Re: These are just shibboleths, aren't they?

[rochvelleth]

I don't buy the etymological arguments concerning data - there are lots of other borrowings into English which have been mangled one way or another compared to whatever language they came from.

Yes, but isn't 'data' generally used as a plural anyway? I mean, I'm pretty sure it was adopted as such, so the change has come later (and internally to English).

I note in passing that I've never heard anyone complaining about "split negatives" in French...

They complain about leaving out the 'ne' though (I hear this is quite common in spoken french). But the 'split negative' construction was always like that, wasn't it? I mean, did it get split secondarily? I'm afraid I don't know very much about the development of French...

... academics are just another language community (or communities), with their own shibboleths

I susepct you're right there. I guess it's just a case of knowing your academics :)

[stephdairy]

Some geographers have data in different datums. ITYSBT.

(S)