Good Reason

It's okay to be wrong. It's not okay to stay wrong.

Category: language (page 13 of 22)

Talk the Talk Twofer: Cave signs

10 March 2010 / / Comments Off on Talk the Talk Twofer: Cave signs

Two scintillating interviews for your enjoyment, all featuring me, and the charming and talented Jamie MacDonald.

First, from the 23 February show: Stroke patients, unable to speak, have re-learned to say words and phrases by singing them instead of speaking.

It’s already been shown that speech and music operate somewhat independently, and some linguists think language might have evolved via music.

Click to listen:

Next, from the 2 March show, a look at cave signs. Why should cave art get all the attention? Researchers from the Uni of Victoria have noticed that some non-representational markings turn up in caves all across Europe. Did they have an agreed-upon meaning? If so, it would mean that the beginnings of a writing system (and the cognition needed to power same) would have happened far earlier than heretofore supposed. When researching this topic, I expected to find a language myth ripe for debunking, but I think it’s pretty solid and the claims are presented fairly modestly.

Click to listen:

I’m on about 5/6ths of the way through the stream. Watch out; it starts playing as soon as the page loads.

Talk the Talk: Retarded

10 February 2010 / / 2 Comments

On this week’s “Talk the Talk”, we discussed the use of the word ‘retarded’. Do you use the r-word? Would you ever describe someone as a ‘retard’?

The issue has come to the fore in recent weeks as Rahm Emanuel, Obama’s chief of staff, used the word to describe Democrats that criticised other Democrats — and subsequently apologised. “Rosa’s Law” has been introduced as a bill to the federal legislature, which would prohibit the use of the r-word in federal documents. And if you’re willing to never use the word again, you can take the the ‘r-word’ pledge.

Or you can just listen to me talking about it on RTRFM.

I’m on about 5/6ths of the way through the stream. Watch out; it starts playing as soon as the page loads.

I’d be critical too.

2 February 2010 / / Comments Off on I’d be critical too.

This headline in the Sydney Morning Herald is causing me some trouble with my word sense disambiguation:

Woman critical after car hits her

This one’s caused problems in headlines before. The Trenton Times (9/2/1982) had a similar headline:

‘Nagging’ wife critical after hammer attack

Isn’t there some other word they could use?

Teenagers getting by on 800 words a day?

24 January 2010 / / 3 Comments

I’m used to hearing people complain about the language of Them Dern Kids, but this rationale is a new one.

Here’s the claim:

800 words won’t get job done

LONDON: A generation of teenagers risks making itself unemployable because its members are using a vocabulary of only about 800 words a day, according to the British government’s first children’s communication tsar.

Communication tsar? Are they sure she’s not a czar?

I wonder what’s causing the supposed paucity of vocabulary? Could it be the Internet and mobile phones?

The teenagers are avoiding using a broad vocabulary and complex words in favour of the abbreviated “teenspeak” of text messages, social networking sites and internet chat rooms.

Thought so.

Jean Gross, the government’s adviser on childhood language development, is planning a national campaign to prevent children failing in the classroom and the workplace because they cannot express themselves.

“Teenagers are spending more time communicating through electronic media and text messaging, which is short and brief,” she said. “We need to help today’s teenagers understand the difference between their textspeak and the language they need to succeed — 800 words will not get you a job.“

Gee, 800 words doesn’t sound like a lot. Or is it? How many words do most people say? Let’s check.

First, keep in mind that the 800 words claim is about daily vocabulary, not total vocabulary. That is, young people are using the same 800 words over and over again in a typical day. I’m not sure if that’s true, but let’s accept it for now. The question is: how many different words do adults employ in a day?

We’re going to use a dialogue corpus to find out. I’m pulling words from Verbmobil-2, a corpus of appointment scheduling dialogues. But we don’t know many words to use until we know how many words someone speaks in a day. This is a scary prospect, laden with assumptions.

I had a read through the corpus and found that I can read about 250 words out loud in a minute. Of course, in a dialogue you’d only be speaking about half the time unless you’re rude, or a lecturer. (Or, like me, both.) So let’s say I’d rip through 7,500 words in an hour. Most of us spend some time alone or watching TV, so I doubt we’d spend the equivalent of 4 full hours of every day talking. But let’s say 30,000 words as an upper boundary. (I admit this is highly speculative. Stay with me.)

Here I’ve listed the number of word types (different words) for various numbers of word tokens (each separate word we say) in the Verbmobil-2 corpus. If you think you’re more laconic or loquacious, you can adjust your expectations accordingly.

Word tokens Word types
10,000 words 814 types
20,000 words 1,080 types
30,000 words 1,342 types
40,000 words 1,510 types

So if you’re an adult on the lower end of the talking scale, you’re going to use about 800 different words, over and over. And even if you quadruple the number of words you say, that still won’t quite double the daily vocabulary. Keep in mind that 40,000 words represents hours and hours of transcripts. The fact is, 800 words is quite a lot. Even if teens only use the same 800 words over and over, that’s certainly not a sign that their vocab is sub-standard. That’s just the way word frequencies fall.

———————-
UPDATE: I’ve just discovered this article in USA Today about a study that saw people wearing tape recorders all day long.

Both sexes say about 16,000 words a day, a study in Science magazine says.
…
He and colleagues analyzed conversations recorded from 1998 to 2004 of 396 students in the USA and Mexico, 210 women and 186 men, ages 18-29. The study examined word count, not vocabulary or word use. Pennebaker says two-thirds of participants spoke 11,000 to 25,000 words a day; the average for both sexes was about 16,000.

So there it is. Sixteen thousand words of dialogue would probably be comprised of under 1,000 word types a day, not too far from 800.
———————-

Let’s take a look at another claim in the article.

Ms Gross said her concerns were supported by research by Tony McEnery, a professor of linguistics at Lancaster University, who found in a study that the top 20 words used by teenagers, including “yeah”, “no” and “but”, account for about one-third of the words used.

Twenty words is not a lot. Is it possible that it could account for a third of the total?

Fortunately, we have frequency statistics for many corpora. If we take a look at the top 1000 words from COLT, the Bergen Corpus of London Teenage Language, we can see that the top 20 words account for 35.6 percent, or about a third. (Some words are excluded from this count, but that just means that the real proportion will be a good deal smaller, which makes the teens seem even more erudite.)

Now we head over to this data from the BNC, or the British National Corpus, a large and wide-ranging collection of spoken and written language. Here, the top 20 words account for around 32 percent of the total, or… about a third.

I decided to run a counter over some works of literature. I tried George Orwell’s 1984. Nobody’s going to accuse Orwell of having a tiny vocabulary. But here the top 20 words account for only 33.7 percent of the total. And for Alice’s Adventures in Wonderland by Lewis Carroll, the top 20 words make up, again, 33.1 percent of the total. Somebody better tell Pseudonym Boys that they’ll never get jobs with that kind of vocabulary.

Gross’s claims sound impressive until you break them down. Most people don’t do this because it’s easier to just accept claims that you already believe. But it’s just another way to complain about young people in a way that’s socially acceptable. It’s a shame people try to enlist linguistic data to confirm their prejudices.
———————-
If you want to hear me say about the same thing on the radio, you can listen to last week’s RTRFM interview. For some reason, I was talking pretty fast. I bet I could have clocked 60,000 words per day at that rate.

I’m on about 5/6ths of the way through the stream. Watch out; it starts playing as soon as the page loads.

Phrase Detectives: Try it out

17 January 2010 / / 1 Comment

One of my research interests is getting Internet volunteers to contribute to linguistic resources. And here’s one such effort, courtesy of Massimo Poesio: it’s Phrase Detectives.

It’s sort of a fun game where you get points for figuring out how pronouns (and other things) refer back to other entities in a text. Of course, the benefit to linguists is that it all goes toward the creation of a 1.2-million-word anaphor resolution corpus. A very worthwhile project.

Please don’t let them quote me.

13 January 2010 / / 1 Comment

Let’s say I gave a very sensible talk about linguistics. And in this very sensible talk, I said that there was a possibility that the spelling of some English words might change as a result of the Internet. And just for fun, let’s say I’m David Crystal. (Oh, come on. Flatter me.)

Anyhow, here’s how the article would pan out once it hit the Sydney Morning Herald.

Internet spells death of English

Traditional spellings could be killed off by the internet within a few decades, a language expert has claimed.

Aaaaugh! Not the spellings!

The advent of blogs and chatrooms meant that for the first time in centuries printed words were widely distributed without having been edited or proofread, said David Crystal, of the University of Wales in Bangor.

As a result, writers could spell words differently and their versions could enter common usage and become accepted by children.

Aaaaugh! Not the children! Won’t someone please, et cetera!

But notice that writers put the most calamitous material at the top of the article. And then by the end of the article, they’ve backed off of all the scary claims, and the whole thing becomes almost sensible.

Professor Crystal told the conference of the International English Language Testing System the internet would not lead to a complete breakdown in spelling rules.

”All that will happen is that one set of conventions will replace another set of conventions,” he said.

But by then, it’s too late because everyone has already fled the house screaming, or are writing angry letters to editors about Kids These Days.

If this article had been about continental drift, the headline would have been ‘Doomed Continents to Collide’.

My advice: When it comes to news articles about language, don’t read headlines if you can help it. They’re written by amphetmine-addled caffeine junkies. Instead, start reading about halfway down.

No Country for Old Men: The coin-toss scene, as seen by a linguist

2 January 2010 / / 3 Comments

By now, everyone must have seen ‘No Country for Old Men’. I’ve only just watched it now — I don’t often have the chance to sit and watch a movie. It’s one of those that keeps coming back to you days later.

The key ingredient in the film is the antagonist Anton Chigurh, a remorseless killer with a Prince Valiant hairdo and an air tank. He’s as omniscient as the next psychotic villain, but he’s not invulnerable; Moss, his quarry, can injure him, and you wonder if that means Moss will be able to turn the tables. Even so, Chigurh has a formidable willingness to dispatch you for the sake of getting your car and continuing his pitiless and emotionless pursuit of Moss, as well as anyone else who crosses his path or even looks at him.

One of the most memorable scenes is the ‘coin toss’, which appears early in the film. It’s a model of how to write film dialogue. At the counter of a gas station, the Proprietor bumbles onto Chigurh’s bad side with a casual question about where he’s come from, and Chigurh won’t let it go. He draws the Proprietor deeper into the conversation and thus deeper into trouble.

Watch:

Unlike the dialogues I study, it’s a fictional conversation, but it lends itself really well to analysis. Two items in my bag of tricks are Conversational Analysis (CA) as elucidated by Sacks, Schegloff, and Jefferson, and game theory, especially Bill Mann’s Dialogue Macrogame Theory (or DMT). CA is concerned with the mechanics of dialogue, particularly the back-and-forth of its parts. Game theory, as I’m using it here, refers to the way people make ‘bids’ to take the dialogue in this or that direction.

From the top:

CHIGURH
How much?

PROPRIETOR
Sixty-nine cent.

CHIGURH
And the gas.

So far, all standard. Chigurh initiates the dialogue with a question, the proprietor answers. This is known as an adjacency pair. We use adjacency pairs habitually; questions lead naturally to answers, comments lead to acknowledgements. It’s the unconscious nature of adjacency pairs that will draw the Proprietor into this tense and dangerous exchange.

PROPRIETOR
Y’all getting any rain up your way?

CHIGURH
What way would that be?

The Proprietor innocently starts a question-answer adjacency pair. But it’s not a question Chighur likes, so he doesn’t answer it. Instead, he takes control by asking a clarification question of his own.

PROPRIETOR
I seen you was from Dallas.

CHIGURH
What business is it of yours where I’m from, friendo?

Uh-oh. Someone has noticed Chigurh’s point of origin, and could rat him out. The Proprietor’s original question is still hanging, unresolved.

PROPRIETOR
I didn’t mean nothin’ by it.

CHIGURH
Didn’t mean nothin.

The Proprietor attempts to repair this situation, but Chigurh won’t have it.

PROPRIETOR
I was just passin’ the time.
If you don’t wanna accept that I don’t know what else I can do for you.

The Proprietor is trying to preclude any further repair attempts. Then:

PROPRIETOR
…Will there be somethin’ else?

CHIGURH
I don’t know. Will there?

People don’t like to close a dialogue down too abruptly, so most dialogues have a ‘pre-closing’ stage, just to make sure nobody has anything else to say. Here, the Proprietor makes a bid to ‘pre-close’ (wouldn’t you?), but instead of meeting the bid with a yes-no answer, Chigurh thwarts the bid with another question. Which the Proprietor needs to address.

PROPRIETOR
Is somethin’ wrong?

CHIGURH
With what?

You give a question, you expect an answer, but Chigurh isn’t cooperating.

PROPRIETOR
With anything?

CHIGURH
Is that what you’re asking me? Is there something wrong with anything?

Chiguhr does it again — he’s not letting the Proprietor take the ‘initiative’ — the first step — anywhere, he’s not resolving any of these adjacency pairs, and he’s using another question to push the dialogue down one more layer. We’re three levels down in this dialogue, which is about as much as people are good at handling. Any deeper and the Proprietor will be lost. So it’s another attempt at pre-closing:

PROPRIETOR
Will there be anything else?

CHIGURH
You already asked me that.

Chigurh gives not another question, but a hostile meta-comment on the dialogue. The Proprietor only has one way out: make a bid to terminate the dialogue proper.

PROPRIETOR
Well…I need to see about closin.

CHIGURH
See about closing.

PROPRIETOR
Yessir.

Bid rejected, using an acknowledgement. Now Chigurh takes control, issuing question after obliquely threatening question.

CHIGURH
What time do you close?

PROPRIETOR
Now. We close now.

A question-answer pair, but Chigurh’s not happy with it. He will decide the level of specificity required.

CHIGURH
Now is not a time. What time do you close.

PROPRIETOR
Generally around dark. At dark.

At last, something resembling a completed adjacency pair. But Chigurh isn’t content to let it rest:

CHIGURH
You don’t know what you’re talking about, do you?

PROPRIETOR
Sir?

CHIGURH
I said you don’t know what you’re talking about.

The Proprietor no longer knows how to play this. He lets Chigurh take all the initiative.

CHIGURH
What time do you go to bed?

PROPRIETOR
Sir?

CHIGURH
You’re a bit deaf, aren’t you? I said what time do you go to bed.

PROPRIETOR
Well…Somewhere around nine-thirty. I’d say around nine-thirty.

CHIGURH
I could come back then.

You don’t want this guy to come back when you’re in bed.

PROPRIETOR
Why would you be comin’ back? We’ll be closed.

CHIGURH
You said that.

It’s the first time in a while that the Proprietor has taken the initiative in this dialogue, but Chigurh shuts him down with another meta-comment about the dialogue itself. Now the Proprietor makes another bid to terminate the dialogue, but Chigurh quashes it with another question.

PROPRIETOR
Well…I got to close now–

CHIGURH
You live in that house out back?

PROPRIETOR
Yes I do.

He knows where you live.

CHIGURH
You’ve lived here all your life?

PROPRIETOR
This was my wife’s father’s place. Originally.

CHIGURH
You married into it.

Chigurh does not attempt to conceal his disdain. The Proprietor must realise he’s in danger, but can’t stop babbling. He’s in this conversation now.

PROPRIETOR
We lived in Temple Texas for many years. Raised a family there. In Temple. We come out here about four years ago.

CHIGURH
You married into it.

Chighur now owns this conversation, and isn’t going to make any concessions.

PROPRIETOR
…If that’s the way you wanna put it.

CHIGURH
I don’t have some way to put it. That’s the way it is.

CHIGURH
…What’s the most you’ve ever lost on a coin toss?

And this takes us to Chigurh’s game, which establishes another part of his character — he’s murderous, but also capricious and arbitrary. The coin toss is probably more interesting for philosophical reasons than for its dialogue, so I’ll stop the analysis there.

It is interesting, however, to note the way Chigurh and the Proprietor discuss the stakes of the game. The Proprietor is no doubt aware of the danger he’s in, but is carefully trying to determine the nature of the danger. They both avoid talking about the stakes of the game directly — the Proprietor, because if he says it, it might happen; Chigurh, because he considers himself an agent of Fate. Discussing it directly would make him responsible, and he’s not; the evil swirling through the film is bigger than this one man.

It’s a rather long scene. One screen-writer says he might have suggested trimming the first part. But you can’t. You can’t just start The Game. First, you have to draw your victim in. Chigurh does this by manipulating the conversation — grabbing the initiative, refusing to resolve any of the Proprietor’s adjacency pairs, and pushing the dialogue down level by level until the situation is inextricable.

Talk the Talk: The language of global warming

10 December 2009 / / Comments Off on Talk the Talk: The language of global warming

A timely interview on RTRFM, this time about the hidden persuaders in language about global warming.

Watch out for that link; it plays immediately, so make sure your speakers are at the right level. As always, I’m on about 5/6ths of the way through.

Talk the Talk: unfriend

26 November 2009 / / 2 Comments

In my latest ‘Talk the Talk‘ interview on RTRFM, I discuss the Oxford Word of the Year for 2009: ‘unfriend‘. Does anyone remember their pick from last year: ‘hypermiling‘? ‘Unfriend’ has a better chance of getting remembered, I’d say.

I also discuss ‘teabagging’, and why I am an amonokerist.

Watch out for that link: it starts playing immediately. I’m on about 5/6ths of the way through.

You know who else liked biscuits?

26 October 2009 / / 4 Comments

Racism is serious, and therefore charges of racism are serious. Sometimes racism gets incorporated into the language in unexpected ways. Like I still can’t believe they have ‘Coon’ cheese in Australia. And my mom remembers a time when brazil nuts were called ‘nigger toes’.

But I don’t think this is one of those times.

What’s in a name? Coles biscuit slammed as racist

The name of a range of Coles chocolate and vanilla biscuits has been labelled racially loaded and a throwback to Nazism by a prominent Brisbane Indigenous leader.

The name of the Coles brand biscuits, Creole Creams, has sparked controversy in the wake of the much-derided Hey Hey It’s Saturday blackface skit earlier this month.

The biscuits, similar to Arnott’s Delta Creams and Oreos, with a chocolate exterior and cream centre.

Sam Watson, the deputy director of the Aboriginal and Torres Strait Islander Studies Unit at the University of Queensland, said the word “Creole”, often used to describe a person of mixed European and Africa ancestry, was a racially loaded term.

“The word Creole comes from a period when people’s humanity was measured by the amount of white blood they had in their bloodstream. This is the same kind of thought that underpinned horrific regimes like the Nazis,” Mr Watson said.

Watson clearly hasn’t spent much time on the net; he’s blown his Godwin on the very first round.

But besides that, I think his argument’s overblown. Two things:

1. Coles isn’t the first to call these things ‘creole cremes’. I found a reference to them on this 2007 post from someone in Melbourne. I’m not sure how widespread this usage is. Anyone?

2. I’ve studied terms of abuse. ‘Creole’ could be used as a term of abuse, but I’m unable to find any examples of people doing so (which I find kind of surprising). On the other hand, ‘Creole’ is used to describe a genre of cooking in Louisiana USA. This page even has a recipe for Creole Cream Cheese, which might go well between two chocolate cookies.

So at this stage, I think this is a frivolous charge of racism. But I don’t know. I thought the biscuits had that name because of the kind of language they invent when talking to other biscuits.

Older posts Newer posts

Meta

© 2024 Good Reason

Theme by Anders Noren — Up ↑