Daily Archives: October 19, 2010


This word seems to me suited to some ogre with a spastic temper; can’t say exactly why. It certainly has two aspects to it: one is the rounded letters and nasal sound of nun, and the other is the sharp éclat of atak, the voiceless stops popping and the k looking like a blasting cap cracking a rock wall. And I confess that those two parts together always make me think of Sister Bernice going berserk.

But while this word does signify something that one may say bursts forth from something very staid, there is no actual explosive action involved. The very staid something it bursts forth from is the byword for slowness and cold: a glacier. But the thing that bursts forth – or rather juts out – is itself the acme of immobility: rock, solid cold hard rock. A nunatak is a peak of rock that juts out of a glacier.

And, like various other names for things associated with the great white north, this word comes to us from Inuktitut – you know, what they speak in Nunavut. And, yes, that nuna in nunatak is the same nuna as in Nunavut: it means “land”. And tak? In current Inuktitut spelling, it would be taq; it means “thing pertaining to”.

And should we spell it nunataq? Only if we want to present it as an Inuktitut word we are inserting in English. But that would stick out unnecessarily – it would be pretentious, in fact. We’ve had this word nunatak in the English language for a century and a quarter, rather longer than they’ve been using q in Inuktitut spelling. It’s also been in Swedish and Danish for as long (Denmark has plenty of nunataks – if you consider Greenland part of Denmark, which officially it is, though it is fairly autonomous). So let none attack our orthography.

Oh, and is it actually “noon attack” rather than “nun attack”? Well, in the original, it would be (but in the original the final stop would be uvular, not velar, and the a‘s would be rather as in father). But while these rocky peaks might seem to attack the noonday sun, we must accept that in the English version the pronunciation has gone the usual English way, and none other.

One or two things about numbers

A colleague has encountered a sentence of the type “This will happen in one-to-two months.” She’s wondering about those hyphens.

And well she should be. They have to go. It’s not optional. Otherwise it’s presenting a type of month with the quality of “one-to-two” in the same way as “two-by-four boards” are boards with the quality of being two inches by four inches. A “moderate-to-severe infection” is “an infection that is moderate to severe”; “one to two months” is not “months that are one to two”.

In cases like this, some people are confused by the use of hyphens in something like “a two-month decline”. But this is not that. In a case like that, the head noun is “decline”, and “two-month” is hyphenated because it is part of a compound modifier. In the case of “one to two months”, “months” is the head noun and the numbers are quantifying it – they are not adjectives, they are quantifiers. That’s another point of confusion some people get into: treating numbers as though they were adjectives. (It doesn’t help that CP Style, presumably for reasons of readability in newspaper columns, prescribes, for instance, “two-million” rather than the standard “two million”.)

“One to two months” is not a set of months with the quality “one to two”; it is “one month to two months” with the first “month” removed. (A similar deletion, but of the final “months”, is seen in “a month or two”, which we don’t write “a- month -or-two”.) We can use a dash to replace “to” in, for instance, “1–2”, but we don’t use dashes (or hyphens) and “to” with number ranges.