Monday, July 11, 2011

Latin lives on

kw: analysis, words, history

As I wrote some text for a planning meeting I was pondering whether to pluralize "curriculum" as "curriculums" or "curricula". I prefer the latter, but I was wondering which my audience would prefer. I did a Google search of both spellings, and it was hands down for the Latin plural, by 48 million to 5.2 million, an 8:1 ratio. I then had two thoughts. One was, I wonder what the history of this is, whether there is a changeover to Anglicized pluralization going on, and whether similar words would have the same history.

When I went to the Google Labs Ngram Viewer, the first pair I tried was "addendums" vs "addenda". This first chart shows the result. The Latin plural is in red, and is the clear winner, for all of the past 208 years I plotted. You can click on these charts for a double size version.

That done, I plotted "curriculums" and "curricula". This chart shows that, while there was some support for the Anglicized plural from 1920 to 1980, it has largely vanished.

Here is a pair for which the drama has yet to play out in full: "referendums" and "referenda". While the Anglicized plural is rapidly gaining ground, and the spell checker here at Blogger tells me it doesn't like "refereda", the Latin plural is still ahead as of 2008. I hope Google Labs updates their database soon. I'd like to see the 2009 and 2010 results.

Also, the plural of "vacuum" has had a strange history. "Vacua" was preferred until 1960, then underwent a precipitous drop in popularity, while "vacuums" had been slowly gaining popularity since 1920 or thereabouts. I expect "vacua" to vanish over the next 20-40 years.

One word of Greek origin has three plurals because many think it is from Latin. The Latin plural of "octopus" is "octopi" and was once preferred, but then was overtaken by "octopuses" after 1920. About 1940 some language purists proposed the proper Greek plural "octopodes", but as you can see, it never caught on. It is effectively defunct. "Octopuses" now reigns.

Finally, there is the strange case of the false Latin plural of "rhinoceros". This word ends in "-os", not "-us", yet I have heard well educated people say "rhinoceri". As this chart shows, this term has very rarely made it into print. "Rhinoceroses" has always been the appropriate plural, because the term was not taken into English from another language, but was compounded of two Greek roots as an English word about 1740 (I ran a different chart to discern this date). This and "octopus" illustrate the mental tricks we have to overcome because we have so many words from both Greek and Latin in our language.

It is fascinating, the products of modern analysis tools when you have mountains of data to work with.

No comments: