This spelling looks peculiar, one would expect can not. Do we know anything about the history of the current spelling?

No, one would not, because it is pronounced as one word.  I have written about it here and here.

languagehat - 26 June 2010 02:45 PM

No, one would not, because it is pronounced as one word.  I have written about it here and here.

In other words cannot should be regarded as an auxiliary verb on a par with can? This would be comparable to the Latin pair of verbs velle (to be willing to) and nolle (to be unwilling to).

OED on cannot:
the ordinary modern way of writing can not

I agree that it is an unusual form: similar forms do not exist for other verbs.

OP Tipping - 26 June 2010 06:26 PM

I agree that it is an unusual form: similar forms do not exist for other verbs.

If we go beyond verbs, is the spelling of another a parallel case?

Forming one word from two is very common in English: there’d be thousands of examples, of which I suppose another is one. What I meant was that there are no other cases that I can think of where -not is used like this.

OED on cannot:
the ordinary modern way of writing can not

Where “modern” refers to the 1880s.  I think we can safely relegate that to the historical shelf.

In other words cannot should be regarded as an auxiliary verb on a par with can?

Yes, I think that’s a useful way to regard it.

As an aside, how do I find out when a particular word was revised by the OED, eg cannot?

It’s easy for third edition entries. The date of update is included at the top of the entry, or where additions have been made (but no changing of the original), the additions are segregated to a section at the end of the entry.

The problem is between first and second edition. Without going back to the original print editions, you cannot tell what, if anything, has been changed.

Short of going back to the old editions, the best clue is the date of citations. If all the citations are nineteenth century or earlier, chances are the entry has not been changed since the first edition. But this can be deceptive—some second edition entries added recent citations or senses, but did not change anything else in the entry.

Wikipedia gives a nice tabular summary of exactly when the various fascicles and supplements have been published, from 1888 to the present.

There’s a better summary on page x of the Compact Edition, giving the publication date of each fascicle:
A-Ant                    January 1884
Anta-Battening    November 1885
and so on.  Too bad it’s not online.

Too bad it’s not online.

The first 5 or so volumes are on the Internet Archive (e.g., Vol. 1, A-B, link). If you click on the HTTP link in the upper left-hand View the Book box, you can download PNG scans of the book. The files are huge; for example the volume linked to is 408MB in compressed TAR format. (I am not sure if it’s available outside the USA.)

That seems like a ridiculous way to do things.  I wonder why no one’s made the material more easily accessible?

I wonder why no one’s made the material more easily accessible?

True, but I think I heard that somebody is trying to digitize it. Then the XML markup can begin. And then the lawsuits.

OP Tipping - 26 June 2010 06:26 PM

I agree that it is an unusual form: similar forms do not exist for other verbs.

However, I think notwithstanding is another example.