A US correspondent, having read The Fight for English, raises this question.
'American spellings: I note with some interest that the OUP house style is to use "-ize" instead of "-ise". I wonder: does this come from Webster (unlikely, I'd have thought!)? Does it suggest an English pre-American practice? Or is it just OUP preference? I guess I shall never know!'
The -ize spelling was preferred by classical scholars, especially in the 16th century, for verbs which came into English from Greek and Latin, and that etymological argument has fostered the use of z ever since. The USA and Canada adopted it from the outset. And the editors of the Oxford English Dictionary opted for it, at the end of the 19th century, partly on etymological grounds (a z is used in Greek and Latin) and partly on phonological grounds (that the letter better reflects the sound). 'In this dictionary', they say, 'the termination is uniformly written -ize'. This influenced Henry Hart, who compiled his 'Rules for Compositors and Readers' at the press in Oxford. He opens his first booklet with a section on spellings, and adopts the -ize spellings used in Murray's dictionary. And Murray, in turn, had been influenced by Dr Johnson, whose Dictionary has agonize, analyze, anatomize, and so on.
So where did the -ise alternatives come from? Some of the words such as baptize) were spelled with both an s and a z from their earliest days in Middle English. The trend to spell all such verbs with s began when verbs came into English with increasing frequency from French, where the suffix was -iser. A verb of this kind borrowed directly from French, it was argued, should be spelled with -ise, to reflect that source. Some felt it important to maintain a spelling link between related words, such as analyse and analyst. And during the 19th century, this usage grew.
The problem, of course, is that it is often unclear whether a verb has come into English from French or from Latin. Confusion led 19th-century printers to try to sort it out, and they did this by imposing a uniform rule for all such verbs where alternatives exist. Hart opted for -ize. But several other publishers - perhaps in an effort to distinguish themselves from Oxford - opted for -ise. They may also have been influenced by the fact that there are fewer exceptions if you go for the -ise rule. Several verbs can only appear in -ise (such as advise, revise, surprise...), and you have to remember what they are.
World usage varies. -ize is the overall preference in North America; -ise in Australia. Usage in the UK is mixed, with -ise beating -ize in a ratio of 3:2. There's a nice discussion of current trends in Pam Peters' Cambridge Guide to English Usage. Usage is certainly changing. Some publishers these days are adopting a more relaxed attitude: they don't mind which authors use, as long as they are consistent. Personally, having had my usage pushed first one way and then the other by publishers over the years, I've given up having a preference!