December 14, 2006

Eggcorn alarm from 2004


While searching for a study reported in BBC News about teenagers and the paucity of their vocabulary, I came across the National Literacy Trust site, which has a database of media items on language-related topics -- many of them, alas, old acquaintances that we have already savaged in Language Log.  Among them was a somewhat fevered story from the Telegraph of 6/8/04 about the "mass dyslexia" induced by... eggcorns.


What set me searching was this very brief piece on BBC News (with a quiz you can take):

Teenagers use just 20 words for a third of their speech, according to a new study by Lancaster University.

And their vocabulary is peppered with slang that can seem - to those old enough to vote and to drink legally - incomprehensible.

Test yourself on the slang used by the teens in the study.

You don't need to write us about this story.  We're on it, and we'll report on it when we have some actual details.

It did occur to me that this could merely be a report on the frequency of the most frequent words in English, in general.  If you look at the Brown Corpus word frequencies and add up the corpus percentages for the top 20 words (listed below), they account for 31% of the words in the corpus.  But that would be ridiculous, and it wouldn't distinguish teenagers from the rest of us, so what would be the point?

The Brown top 20: the, of, and, to, a, in, that, is, was, he, for, it, with, as, his, on, be, at, by, I

While I was thrashing about on the net, i came across the National Literacy Trust and this piece from 2004, which I'll divide into two pieces:

Howlers of modern English usage

As part of compiling the 11th edition of the Concise Oxford Dictionary, researchers discovered an increasing confusion over simple words and phrases. One in five believes we should 'tow' the line instead of 'toe' the line. A further 10% 'pour' over a book when we should 'pore' over it. The alarming increase in 'mass dyslexia' was picked up by the dictionary's 100 researchers worldwide on the look out for new words. As they searched the Oxford English Corpus, a database of 400 million written words, they discovered that while spelling remained reasonably strong, more and more writers were mixing up like-sounding words and phrases. 'They are not so much spelling the words wrongly, as using the wrong words,' Angus Stevenson, the dictionary's co-editor said.

Yes, they're eggcorns, famous ones, and only 8 months before these errors got their name, right here on Language Log.

I'm suspicious about the things-are-getting-worse tone of the piece, though: "increasing confusion", "alarming increase", "more and more writers".  Did the COD researchers actually have comparative data showing that more people were writing "tow the line" and "pour over a book" when the COD11 data were collected than they did, say, twenty years before?  (It's not particularly hard to find such things from fifty years ago, though I have no idea of their frequencies then.)  Or were they merely alarmed at the 20% figure for toe >> tow and the 10% figure for pore >> pour?  I wouldn't be at all surprised if the percentages for these items were rising, since (like other frequent eggcorns) they make a lot of sense to people and so would be inclined to spread -- but is there any evidence on the question?

Now, the rest of the story:

His team believes that the chief explanation was the use of the computer spell check, which does not spot errors of meaning. He also thought that the explosion of the numbers of people writing, mainly due to the internet, meant that more errors were bound to creep in. Whether such mistakes will, in time, spill over into more formal types of writing is yet to be seen. The question is: does it matter if in a generation's time people are writing about 'pouring over magazines' or 'towing the line'?

The findings come after a campaign, backed by Bill Cosby, the American actor, began to stop British children from speaking patois in class. A south London school is piloting a scheme to ban slang, often based on the creole spoken in the West Indies, because it is thought to contribute to the educational failure of black pupils.

(Telegraph, 8 June 2004)

The explanation in terms of spell checking programs isn't entirely clear to me.  The assumption seems to be that people used to learn to spell the hard way, by endless drill and correction by teachers and editors, until they got it right, but now they rely on spell checkers to fix things for them, so that errors like toe >> tow no longer get caught.  This isn't entirely implausible, but is it true?  Did students used to get a lot of correction on such items that would lead them to change their practice?  Do they not get that now?

The question of whether, ceteris paribus, spelling is getting worse is not an easy one to answer.  There are a number of corrections that have to be made before things are roughly equal -- to take into account the opening up of higher education to a much wider segment of the population after World War II, for example, and to adjust for the much wider appearance of informal writing in public in recent years, on the net in particular.  The materials that are being compared have to be similar in character.

There's certainly a widespread PERCEPTION that things are sliding downhill.  But that doesn't make it so.

In any case, as the Telegraph piece asks, does it really matter if toe >> tow and pore >> pour go most of the way to completion?  The expressive resources of the language will not have narrowed, and people will still understand one another (so long as they're behaving cooperatively).  Things change in a tiny way.  That's no Sign of the Apocalypse, just business as usual.

zwicky at-sign csli period stanford period edu

Posted by Arnold Zwicky at December 14, 2006 08:24 PM