Saturday, September 06, 2003
Another latecomer to the debate on language rights

Cosma Shalizi has also commented on my recent series of posts on language rights. Writing the follow-up post just keeps getting harder and harder, and I fear I'm going to have to turn it into a another multi-part post. Also, my hit counter leads me to suspect that several authors of the book I started by reviewing - Language Rights and Political Theory - have hit my blog. If you're one of them, hi!

Cosma pulls out the heavy guns. Yes, ladies and gentlemen, it's Popper time. I will get to that soon in a follow-up post. I just got back from a birthday party - hi, Peter! - and am a bit too tipsy to try it now.

But there is one thing I want to draw attention to. Cosma is "a little disappointed that [I] didn't at least mention a certain recently-influential tradition, which still has adherents, that anticipated [me] in certain aspects."

Cosma is speaking of the big man himself - Karl Marx. Now, lemme see. I talked at length about Lev Vygotsky - famous Marxist psychologist - and quoted Marx himself, not once, but twice without attribution. I linked to a Norman Geras article which does nothing but talk about Marxism. "Self-development" itself is also something of a Marxist tagline.

In short folks, it's time I confessed. Cosma has smoked me out. I was advancing a theory which is Marxist through and through, drawing heavily on Marxist literature and the Marxist tradition. I was being snarky by carefully and intentionally not using his name. I expected someone to point this out, but Cosma is the first to call me on it. The prize in this unannounced competion is 1,000 Laotian Kip in cash - acquired during my vacation two years ago - collectable by personally coming to Belgium to pick it up.

Friday, September 05, 2003
Things Technorati misses

I have missed linking to one of the better philosophy blogs out there - Wäldchen vom Philosophenweg. Worse, it's written by a regular commentor here and has a post on my mammoth posting on language rights.

I can only apologise. I rely to much on link services to tell me who is linking to me. Folks, if you want me to read your comment or link to you, you have but to drop me a line.

Stupid things people say about language

This article has me shaking my head:

Brand names create global language

Brand names have become so abundant that in France they account for two out of every five words an average person knows, according to a study being carried out by French branding company Nomen.

Shocking as that may sound, Nomen Chief Executive Marcel Botton says the trend is creating a new international language that helps people communicate in foreign countries.

" You hear people abroad asking for well-known brands of food or drink when they don't know the word in the foreign language. It's irritating for Coca-Cola when rival products are treated as the same thing but it makes people's lives easier. "-Nomen Chief Executive Marcel Botton

"The distinction between brand names and ordinary words is becoming quite blurred," Botton told Reuters.

"This is bad news for companies that have invested a lot of money in branding a product, but for the general public I see advantages. Brand names are more international than words and they are creating a new Esperanto, which I rather like."

Esperanto, an artificial international language invented in 1887 as a second language everyone could learn, never took off.

Learning brand names, on the other hand, is subconscious, as names seep into our brains through advertisements and shopping.

English speakers use trademarks like Frisbee, Hoover and Walkman as ordinary words, causing some to spread abroad. France uses "Kleenex" for tissue and "Scotch" for adhesive tape.

"You hear people abroad asking for well-known brands of food or drink when they don't know the word in the foreign language. It's irritating for Coca-Cola when rival products are treated as the same thing but it makes people's lives easier," Botton said.

Botton, who created such company names as Vivendi, Wanadoo, Arcelor and Vinci, began testing people in August to see how many brands they recognised and how many dictionary words they knew.

"It crossed my mind this would be an interesting study, as it seemed people knew more and more brands and fewer words."

Botton's team chopped up the French dictionary -- which lists around 100,000 words -- into bitesize chunks which were read out to different people within a test group. A list of 20,000 brand names was also split into chunks and read out.

While the study is not complete, the results so far suggest the average French person knows some 3,000 words and is familiar with around 2,000 brand names on top of that -- suggesting that 40 percent of total vocabulary consists of brand names.

Botton said there were enough possible permutations of pronounceable one-, two- and three-syllable words to cover several billion new brand names.

I presume that what Botton has discovered is that average French people can tell you what roughly 3% of the words in the Petit Robert mean. That seems a little low to me - I would have expected a figure closer to 10% - but it's probably the right order of magnitude. That part is not too surprising. The thing is, that doesn't mean you only need 3000 words of French to get by in life. Explaining this will require a bit of basic lexicology.

First, dictionaries don't list words, they list lexemes. These are not the same things. To be is a lexeme. Be, am, 'm, is, are, was and were are words. In most languages, fluent speakers have to understand the meaning of several words for each lexeme. Second, a surprising percentage of the words in everyday texts are proper nouns of various kinds that don't appear in dictionaries. But most importantly, a count of lexemes is not a real measure of the amount of linguistic information a fluent speaker has to possess.

Consider the sentence "It was raining hard." In French, you could translate that as "Il pleuvait fort", which literally means "It rained strongly." One of the pieces of information you have to have in your head about the words rain and hard in English is that when there is an intense rain, it rains hard and not *"It rains strong." ("*" is a symbol used in linguistics to designate an intentionally incorrect example.) So, there are only two words rain and hard, but not only do you have to know what rain and hard mean, you have to know what hard means when it refers to rain.

This phenomena is called either collocation or lexical functions. Since I started out in the Meaning<->Text school of linguistics, I tend to prefer the term lexical functions, reserving collocation for any pair of lexemes that frequently occur together, for whatever reason. Since I mostly do statistical language processing, the distinction is important. By taking lexical functions into account, the number of meaningful units a native speaker has to know explodes by a factor of anywhere from 10 to 50.

To this, we have to also add a certain amount of phrasal information. Consider the sentence "He does what he wants." To do has its most common meaning in that sentence. Now, consider the differences in meaning to do takes in these three sentences:

     What can I do for you?
     What did you do to my car?
     I want to do her.

To do for X, to do to X and to do X all have completely different meanings, and competent native speakers have to distinguish between them. This same phenomena occurs in French. Faire à X doesn't mean the same thing as faire pour X. These are sometimes called phrasal verbs, but phrasal verbs covers another class of phenomena in English - the Germanic separable verbs. These often don't appear in dictionaries, although most dictionaries make an effort to cover them nowadays. An example of a separable verb in English is to take off, which means something that can't be deduced from the meaning of to take and off. Standard French has phrasal verbs but doesn't have separable verbs, although Canadian French sometimes does:

- T'as-tu vu mes clés?
- Oui, quand je suis allé au bureau, je les ai pris avec.

Lastly, words have very different meanings depending on the register of the language. The word to book means something very different when you're watching Hawaii 5-0 - "Book'em Dano, murder one." - than when you're an accountant for a manufacturing firm - "Book the new orders before the end of the month so they show up as revenue." High frequency words tend of have multiple distinguishable meanings. Although these meanings may (or may not) be related in some way, they can not simply be deduced from a single definition. To know the word to book, you have to know that in means something different in the law enforcement register than in the accounting register - there is no single meaning of to book which by itself tells you what each one means.

Proper nouns - like brand names - are not like common class words or even technical terms. They rarely have any irregular morphology, so they only have the most standard inflected forms. They usually don't have any non-trivial lexical functions. They usually don't change meaning based on the prepositions or phrasal context that you find them in. And, they usually don't vary across registers.

So, it seems very silly to suggest that brand names represent 40% of the words people know. I doubt very strongly that more than the tiniest part of people's lexical knowledge is bound up in brand names. Even sillier is the idea that brand names are forming the core of a new Esperanto. I mean, come on! How many brand names are verbs? Or adjectives? Or adverbs? I mitsubishily kleenex the nike coke?

It is true, however that brand names sometimes become common nouns, much to the annoyance of their trademark holders. Kleenex, for instance, is not happy having their brand name in use as a common word. And, it's true that in much of the world, when you order a coca, you may well get Pepsi. As a linguist, this strikes me as remarkably normal behaviour for people. Besides, I rather like knowing that the more successful a brand is, the more likely people will take the brand and use it as a generic term, destroying its value. It's something I like to encourage.

Thursday, September 04, 2003
If you read H. P. Lovecraft...

...just follow the link. Via Three-Toed Sloth.

Wednesday, September 03, 2003
Enetation, once again, fails to earn the money I don't pay them

Comments are on the blink again. Yet another reason to move to MT.

A Link from An Unenviable Situation

I wanted to mention that D. Ghirlandaio at An Unenviable Situation has also linked to my series of posts on language rights. I bring this up now, because there are no permalinks on his blog and the post may drop off the page before I get to responding. He makes an important point that I would like to clarify. I intend to get into a little more depth in the follow-up post.

I can make a point here that applies also to the question of language politics: Languages can only die a natural death. Invasion and conquest are as natural as the search for food, but we are at the point in our history that in the cause of our morality we defend an ethos of artifcial preservation and museification. My civil morality makes me defend this process, but not without regret. The author of "Pedantry" is too willing for my taste, to replace living culture with its simulacrum, but he is trying to come to terms with just what that culture is.

He's got me dead to rights on the last point: This is about trying to come to terms with what culture is. However, the cultural-historical activity school in education is particular in the kind of distinctions it makes between "natural" and "artificial" in measuring a person's abilities. It explicitly and emphatically makes no such distinction. I deny, for example, that a child's arithmetic skills with pencil and paper are "natural" and arithmetic skills with a calculator are "artificial."

Since I'm extending cultural-historical activity theory to political philosophy, I am also extending the unwillingness to deride one process as artificial and another as natural. Just as it would be wrong for me to label the death of Occitan as "artificial", I won't label using legal mechanisms to preserve French in Canada as "artificial" either. Nor do I class the mechanisms that preserve English in the US and French in France as "natural."

He's right that invasion and conquest are very natural behaviours, just like the search for food. However, we still develop theories of right and wrong that make it wrong to steal food. There is nothing unnatural about slavery or genocide either, but I don't think many people would want to subscribe to a normative political theory that labelled them as "right."

The point is not that one kind of language death is natural and the other artificial. I don't think I ever made any such distinction. I've proposed that language death is always unjust, but that injustice is not always avoidable or remediable.

Update: Seth Edenbaum does have permalinks. The relevant post is here. He also explains that I've missed the point in the comments (which are hopefully still available) and here if not.

New venue

I am now blogging part-time at A Fistful of Euros - a group blog on European affaris, events and, well, stuff in general. Go, take a look. There are a number of excellent Eurobloggers who have joined the FFOE collective: David Weman, Matthew Turner, Jurjen Smies, Tobias Schwarz, Iain J Coleman, and Nick Barlow. Add it to your bookmarks.

I will continue this blog, but I expect to post fairly little on the news here and stick to more involved posts on Pedantry. In fact, I'm preparing to move this whole operation to MT this month. My posts from my grandfather's memoirs will continue, and I intend to do a follow-up on my language policy posts (see below) this week. Things have been a bit slow here on the blog the last few days, primarily because work has picked up a lot and I'm hunting for an apartment in Brussels, but I'm expecting things to start picking up very soon.

Sunday, August 31, 2003
Aaron McGruder says what I've been thinking