-ize / -ise

The Random-ise Project

I've been doing a little investigation into the use of -ize suffix verbs and their spelling (-ize or -ise) in British English.
But I'm also interested in the development of the use of -ize suffix verbs (eg realize/realise) compared with the use of French-based -ise verbs (eg surprise). And in the extent to which there was consistency in the spelling of these verbs, or a lack of it, in early publishing.

Magazines and periodicals

These now have their own start page here

Method

For books, I first check the text at Project Gutenberg. This is an HTML file and very easy to search (especially with a little program I wrote to do the donkey work). Having found all the instances of -ize and -ise verbs at Project Gutenberg, I try and find a British-published edition at Google Books as near as possible to the original publishing date and see what spellings they used.
Searching digitised versions of old books at Google Books or elsewhere is not always so straightforward, as some parts of the text have been corrupted, due to the condition of the original book that has been digitised. Or words are split over a line-break, and don't show up in search. And in older books, verbs like realized are sometimes printed as realiz'd, just to make life interesting.
As an example, for Milton's Paradise Lost, Google Books comes up with all eight instances of advise found in the Project Gutenberg text, but searching the PDF format of the facsimile at Archive.org brings up none. I wondered if this might be due to the practice of using of an f-like letter to represent s in early publishing, and sure enough, searching for advife and surprife is more successful.
What I have learned is that with digitised books, there seem to be two digital versions: the (photographic?) one we see, and a background plain text version, and it seems to depend on how this latter has been treated as to how successful searching is. In one case at least I have found texts shown with s endings in the facsimile showing up with a z in search.
I deal with magazines rather differently. Here it is difficult to find HTML texts at Project Gutenberg, except for single issues, which are very time consuming to search. Most magazines were republished in yearly or more likely six-monthly volumes. These I try and find at Google Books, but you can only do exact whole word searches at Google Books, so I have a set group of ten of the most common -ize verbs, and I search for them and their variants, both with a z and with an s.

Conclusions so far

  • The very first -ize suffix verbs to come into English, verbs like authorise and recognise, took the Old English ending -isen, as the letter Z hardly existed in Old English. Once the letter Z and its pronunciation had become established, however, these verbs began to be spelt with a Z, and by the end of the 14th century, the Z spelling was more or less standard.
  • These -ize suffix verbs were not used very much until the 18th century. On the other hand, use of the limited number of French -ise verbs was very popular. They were occasionally also spelt with a Z.
  • From at least the end of the sixteenth century,as well as adopting -ize suffix verbs from Latin and Greek, often via French, the -ize suffix was being added to existing native English nouns.
  • There was a lot of inconsistency in spelling of both -ize suffix verbs and French -ise verbs right from the beginning, and well into the middle of the nineteenth century.
  • In the eighteenth century, the use of -ize verbs increased substantially, as is evident, for example, Laurence Sterne's Tristram Shandy, 1759.
  • For some reason, some British publishers started to spell -ize verbs with an s in the first half of the nineteenth century. The move from z to s happened very rapidly. All six of Jane Austen's novels were originally published (1811-1818) with -ize endings. When Richard Bentley started republishing them in 1832-1833, in the 'Standard Novels' series, they had nearly all acquired -ise endings.
  • Conventional wisdom puts this change by British publishers down to a mistaken analogy with French -ise verbs, but so far, I've been unable to find any contemporary discussion of the topic. My gut feeling is that it had a lot more to do with consistency. Once publishers changed to s, the levels of consistency improved greatly.

2 comments:

Peter Fennell said...

Dear Will,
This is a really class piece of research.
I have found it because I am proof reading a book with a mix of ize and ise.
Many seem quick to settle on a right position (with OUP) on superficial justification and I am grateful for your industry in checking so much primary source material. As you point out most of these words are modern inventions by analogy. You can hear that in their sound. Organization and Standardization sound very un-Anglo-Saxon!
I love etymology and Douglas Harper's online dictionary. Thank goodness modernisation of spelling ain't goin' to happen (as you say) or many of the archaeological traces of a word's origins would be dusted over. But in this debate meaning is not at stake. Or archaeology. It is obvious when a word has been "ise-ised". Exorcise being the notable exception. I would have guessed it involved cutting something out. ironic that it is the one word that is not zed-ised!
So what is the etymological case for z? How many of these words really came directly from Greek or Latin? Rather than through French? What you have shown is that most of them were invented by analogy later. So there is some rather unauthentic post-rationalisation in the etymology case.
Here is an etymological objection. Monopolise, monopolies, monopolist, monopolistic. The 'iste' habit no doubt we got from French. But like many of our habits it might have come direct from Latin -ista, or Greek -istes. Either way it seems to me there is a family of related endings. All in s. 'ism' is another. They look related and if they are that is a possibly deeper etymological truth that the z spelling hides. In Greek some operation has changed sigma to zeta. No doubt an operation that in Sanskrit would be part of the rules of Sandhi but which in other languages (including Greek) the etymological traces of which are lost. At this distance it looks like an anomaly and the more -ists and -isms we make the more anomalous it looks. It is possible that a deeper etymological dive into the rules of Sandhi would give a clue to how or why it occurred in Greek but whether the same anomaly should be re-manufactured in English seems debatable.
There is a rule in science that you don't express the dimension of a thing in units more accurate than you can measure it. Bricks are too nobbly to measure a wall and say it is 5500.3mm long. I feel the z spelling starts to contravene something like that rule. By analogy don't convey a clean history where there is not one?
English is good at adopting. We like British Exceptionalism even though we picked some of it up from French - historical truth tends to be messy that way but we can cope with that. We have an Unwritten Constitution and a Common Law that are both made up of the book of all our past mistakes. We find that authentic and more reliable than a cleaned-up Roman or Napoleonic Code. We frequently dip in and find new interpretations that a Code cannot offer. So we are suspicious of codification.
I can see why the new studies in Greek caused z to be adopted by publishers. There was also a whole counter movement in C19 toward Anglo-Saxon origins of which Pugin was a part in architecture so there is a Classical v Gothic aspect. Our gothic cathedrals keep this choice always before us and our much-loved Houses of Parliament were built in the neo-gothic conception as a result. America's White House and Congress of course both Classical.
At this point I feel happy to keep both options in British English and favour the s.
I would advocate authors use z when they want to convey:
- officialdom bureaucracy or irony
- modernization or americanization
- scholastic attitudes especially the professional belief that science has the single source of truth
Otherwise if in doubt or no axe to grind use s.
Publishers would use s if agnostic. Or z or s to suit the tone of their author.
Best wishes. Thanks again.
Peter Fennell
peter_at_fennell.com

Peter Fennell said...
This comment has been removed by the author.