Language Log

These ones (cont.)

My mail about these ones and related matters is piling up alarmingly.  The big news is that the expression is indeed regionally distributed: my U.S. correspondents are mostly dubious about it, but my U.K. correspondents find it unremarkable (and were consequently astounded by my judgment that it was non-standard).  And, not surprisingly, several of these British speakers report that these ones and plain these are not equivalent for them.


First in was Nicholas Widdows, who said that he


but went on to consult the British National Corpus, where he found examples suggesting that these ones has a use that isn't easily available to these, namely to pick out one instance as a representative of a type:


He continues:


And ends with an


Not so crazy: Alex Boulton tells me that some French teachers -- Boulton is at the University of Nancy -- tend to think that ones is an impossible form, period.  This is what you'd think if you appealed to "logic" and also subscribed to the idea that the one of this one is the numeral one -- in which case a plural ones would be "illogical".

However, the ONE of this one and these ones is not the numeral, but an indefinite pronoun, which serves as the head in NPs; the numeral, in contrast, functions (like other numerals) most commonly as a determiner (though, again like other numerals, it can serve as head in some constructions, for example the Numeral of NP construction in one of the dogs).  The numeral is the historical source for the indefinite pronoun, and for the generic personal pronoun (as in One never knows), and for that matter the indefinite article a(n), but these lexical items have all gone their own ways long ago, and each has its own syntax and semantics.  Etymology is not destiny.

Several correspondents have suggested treating these ones as parallel to these two, but I think that the real parallel is between these ones and this one, both of them demonstrative + indefinite pronoun.

[Digression on these two, which has two points of interest.  One point is that it shows that sometimes determiners can be layered, as in these two ideas (cf. the/your two ideas and the/your one idea here; there are other types of layered determiners).  Another point is that the two of these two fuses the determiner two with some indefinite head element -- much like ones, in fact -- so that two in some sense realizes both parts, much as the professor's in I like your idea, but I like the professor's even more realizes both the determiner the professor's and an indefinite head -- much like one -- in a single constituent.  (The fusion idea comes from CGEL, ultimately from the work of Michael Wescoat.)]

In any case, there's nothing wrong with ones in general.  Things like the blue ones and which ones and the ones (with a postmodifier, as in the ones from Chicago) and so on are all fine.

[Another side issue: correspondent Empty Pockets reports having been taught never to use these "as a noun" (that is, as a pronoun), but only "as an adjective" (that is, as a determiner), and suggests that overzealous application of this "rule" might be behind the use of these ones instead of these.  I'm familiar with this proscription, posted about it to ADS-L back in 2003, and intend to re-work that posting for Language Log.  But what's important here is that the proscription is general, applying to anaphoric uses of all the free-standing demonstratives: this, that, these, and those.  (The justification offered for this proscription is that all such uses are "vague".  Yes, this is hogwash; you don't need to write me about the deficiencies of the idea.).  There would be no reason to "fix" only one of these four by supplying one(s), and anyway, in the examples we started with it was deictic, not anaphoric, uses that were at issue.]

Back to the geography.  These ones seems widespread and in no way notable in the U.K.  The BNC, according to Alexander Boulton, has 87 occurrences of these ones and 60 of those ones, which is not a lot in 100 million words.  But the reports from native speakers suggest that the low numbers in the corpus merely reflect the rarity of the situations that would call for these expressions; when they're useful, British speakers use them.

Andrew Cave reports (from Brisbane) that it's also common usage in Australia.  Meanwhile, Dan Asimov started noticing it around him when he moved this summer from the U.S. to Vancouver Island, Canada.  He posted about it on the Wikipedia "talk" page on Canadian English, but the follow-ups to his query don't tell us much about Canadian usage on this point.  Now Fiona Hanington writes -- also from British Columbia -- to say that it sounds completely natural to her.

In the U.S., I have a report from Patrick Whittle, saying it was reasonably common in central Kentucky, where he grew up.  And from Dave Kathman, who grew up in suburban Chicago, saying that it's perfectly OK in his idiolect.  As a card-carrying linguist, Kathman went on to say something about how he uses it:


This account is similar to the one from Nicholas Widdows above, and to a suggestion from Andrew Clegg that the 1 variant, with ones, is more likely when there's an implied contrast.

It's not surprising that people with both the 0 and 1 variants should subtly differentiate them -- as I repeat every week or so here on Language Log, variation is very rarely truly free -- or that the 1 variant, with an explicit head pronoun, should be used for contrast.

On the geography and standardness fronts, things are not entirely clear in North America.  A contributor to the Wikipedia talk page ("the user formerly known as JackLumber") points out that


But then we have Dave Kathman's judgments, and the practice of a number of Canadians, at least in British Columbia.  It's possible that in North America these/those ones is a variant in the gray area between standard and non-standard -- fully acceptable to educated middle-class speakers in some areas, but not fully acceptable, though not actually stigmatized, to such people in other areas.  I know of cases like this.

For example, we had a discussion on ADS-L back in May about the sentence-initial discourse connective (serving as an additive adverbial) as well, as in


(Please note how I've delimited the as well under consideration here.  There are several other uses of as well, with different syntax, semantics, and sociolinguistic status.  What I'm reporting here doesn't carry over to the others.)

Garner's Modern American Usage (p. 71) identifies sentence-initial linking as well 'also' as a Canadianism:


The facts seem to be that this discourse connective as well occurs with some frequency in both the U.S. and Canada, and that in Canada it is subject to no stigma, while in the U.S. it is viewed by some commentators as at least informal (and by some as unacceptable) and occurs infrequently in "good writing".  Maybe these/those ones is roughly like this as well.

This would probably be a good place to halt posting on ones.  I got into the topic through a search for literature on the regional and social distribution of these/those ones and on its history, but I've found nothing.  I wasn't proposing to investigate these topics from scratch myself, and if I were, collecting e-mail reports from individual speakers about their usage would not be an appropriate research strategy, nor would just googling up examples.  The most you can get from such sources is a sense of what you might look at in a careful study, and a careful study is a very big project.

But before I bow out, here's an intriguing report from Nick Baker on yet another side topic:


Posted by Arnold Zwicky at October 5, 2007 03:34 PM