Arnold Zwicky's Blog

(Or, for that matter, why not onety-one, onety-two?)

These are questions that kids acquiring English frequently ask, as Ruthie does in this One Big Happy strip:

(#1) Kids are pattern-seeking organisms, and when confronted with a deviation from a pattern, are inclined either to silently, unconsciously, eliminate the deviation in favor of the pattern; or to explicitly lodge a complaint, as Ruthie does here

Ruthie asks a why-question, and there is an answer to it — a complex, fascinating answer supplied briefly for the general reader by Arika Okrent in several places (see below) — but it’s of no use to Ruthie, because that answer is about the history of English, while she wants to know what sense it makes for English to work this way now, and why don’t we fix it.

Alas for Ruthie, the only fair answer to her question is that English just is as it is, and it doesn’t really make sense. Bite the bullet, kid.

The patterns. Kids discover the patterns in English number names in the process of learning to count — leaning to reproduce the sequence of English cardinals (one, two three … ten eleven twelve … sixteen seventeen … twenty twenty-one twenty-two … sixty sixty-one sixty-two …) and to use this sequence to assess the size of collections of things. At the beginning, this is all sheer rote memorization; there are no patterns to discern.

But by the time you’ve learned to count to 20, you will probably have seen a pattern in the teens, that eventually they settle into words of the form A:

N1 + teen (where N1 is one of the ones ordinals one … nine)

(with some pronunciation twiddles in thirteen (for threeteen) and fiftteen (for fiveteen)). At this point, you realize, unconsciously or explicitly, that eleven and twelve are just bizarre; they should be oneteen and twoteen. And a fair number of kids then just fix things to make it so.

For these kids, there’s a brief period where they produce eleven and twelve, just chugging along with raw imitation of what they hear. Until they get the information that shows them the pattern, in which case they start producing oneteen and twoteen.

Linguist parents rejoice at this triumph of pattern-finding; everyone else is alarmed that their kids seem to be regressing. This is a familiar pattern in language acquisition; its most common manifestation is in the acquisition of irregular inflectional morphology: a kid at first gets irregular forms like the PST V broke correct; then regularizes them (to breaked, in this example). Linguists rejoice, normals freak out.

But there’s no reason for alarm. Kids are almost always swayed by the weight of the evidence from other people’s speech that the irregular variant (broke, oneteen) is, regrettably, the one to use, and shift back to it, all on their own. Life goes on.

More patterns. (Advanced, optional, material.) As kids get further along in the sequence of ordinals, they will pick up on the oddity of the form A pattern. There are two parts to this.

The first part is the names of the N10 ordinals (the tens ordinals). From twenty through ninety (with pronunciation twiddles for twenty, thirty, and fifty), these are of form B:

N1 + ty

The second part is the composition of N10 words with N1 words to yield the sequence of ordinals from ten through ninety-nine; from twenty through ninety-nine, these are of form C:

N10 + N1 = (N1 + ty) + N1

Now you’ll see the oddity of form A; the expected ordinals for the first N10, ten, would be

ten + N1 = (one + ty ) + N1 : onety one ’11’ … onety nine ’19’

So oneteen is pre-empted by the special eleven, but in fact oneteen would itself be a special pre-emption of the expected onety one.

Oneteen and twoteen are fairly common creations of kids acquiring English. I believe that onety has occasionally been created, but  it seems that the Form A words, N1 + teen, are learned early and are used frequently, so they’re pretty much locked in by the time the kid gets clear evidence of the Form C pattern.

Pattern conflict. The Form C pattern, with N10 + N1 (larger unit before smaller one), is the general one for English ordinal names It’s the pattern that extends to still higher-level ordinals:

N100 + N10 + N1 (four hundred sixty eight)

N1000 + N100  + N10 + N1 (two thousand three hundred eighty three)

and so on

The pattern for the teens, Pattern A, with N1 + teen, has the smaller unit before the  larger, and might be read as ‘N1 on ten’. (This contrary pattern is the one German also uses for the teens — achtzehn ‘eighteen’ — and then extends, via a conjunction construction, through to 99: neun und achtzig ‘eighty-nine’, literally ‘nine and eighty’.)

But why? The answer to this question wouldn’t actually be of any use to Ruthie, but it’s intellectually intriguing in its own right. Enter Arika Okrent, with her 2021 book Highly Irregular:

(#2) See the chapter “Six of One, Half a Twoteen of the Other: Why Is It Eleven, Twelve Instead of Oneteen, Twoteen?”, p. 75-78

(The chapter developed from her Mental Floss column of 2/24/16, “Why Is It ‘Eleven, Twelve’ Instead of ‘Oneteen, Twoteen’?”)

We learn that eleven and twelve were probably originally words translatable as ‘one left’ (after ten) and ‘two left’, and that these ordinals were especially useful in everyday life (think about dozens), so that their utility and frequency tended to protect them against the forces of regularization. Okrent on p. 78 of the book:

The words we needed earliest and used the most frequently are usually the most irregular. Eleven and twelve were about as high as we needed to go at some point. Their weirdness is a sign of their importance.

 

 

This entry was posted on March 3, 2023 at 2:42 pm and is filed under Child language, Linguistics in the comics, Numbers. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.