Friday, August 08, 2003
Language Quizzes and Chi-Squared Tables
Via Language Hat, I found this collection of language proficiency tests at Transparent Language, a seller of language self-study software. Having nothing better to do today, I took several of their quizzes. Here are the scores I got:
These are pretty mixed results. My French was once better than that, but I hardly ever use it here. But I was shocked at my German and Dutch scores. You see, I haven't spoken German at all since the mid-90's, and it was never my most fluent second language. I did grow up hearing various kinds of German around me, so it never seemed completely foreign, but I only spent two semesters in German classes in my life. My Dutch score, though, was really disappointing. A lot of the questions revolve around the order of verbs at the ends of sentences, which is a big problem for me because I always want to put the words in German order, and that is wrong 90% of the time in Dutch. Also, I have to admit to not having a good enough vocabulary in Dutch.
I was also surprised by my strong Spanish score, since my Spanish consists of four credit hours in 1990 and nothing since. I would have been mortified if I had scored better in Spanish than in Dutch after taking the equivalent of 14 semester-hours and living in a Dutch-speaking country for two years. Still, as the website itself points out, one should not place too much stock in these little quizzes.
I was gratified to have any sort of score at all in Chinese. I haven't spoken a word of the language in two years, and I have to take the placement test here in Leuven in just under a month. However, the relatively low score forced me to ask myself how many of my right answers could be attriuted to chance. The tests are structured as 50 multiple choice questions with four answers in each. Therefore, randomly choosing your answers should give you an average 25% right, or 37.5 out of 150 points. I only had 63 points on the Chinese test, so I asked myself, what are the odds of getting a score like that randomly? How much do my results actually reflect any linguistic knowledge and how confident can I be that chance isn't reason for my scores?
So I consulted a Chi-squared table I found on the web. Now, you would expect random selection of answers to give you 37.5 correct. (1/4 * 150) So, my Chinese score of 63 leads to a p value of 23.12. There is only one degree of freedom in this test, so we look up 23.12 in the top row of the chi-squared table and we get less than 0.0005.
So, we can conclude with fair confidence that I actually do know some Chinese, and all those Chinese classes haven't been completely for nothing.
In order to test this hypothesis, I decided to take the Irish test. I told myself, how much Irish have I studied? There was a lesson on Celtic languages in my historical linguistics class in '92, and there was that day in '95 that I spent in suburban Dublin, drinking cider in a pub and learning Irish from a "Teach yourself..." book with the help of a bunch of drunks. But hell, I barely even remember that day. I wasn't even conscious for most of it. I have absolutely no memory of any Irish I might have learned, and I haven't looked at it since. I have no Irish skills whatsoever.
So, I took the Irish test, figuring I would get a random result. Instead, I scored 51 points - 12 less than I got in Chinese, which I studied on and off for years. The p value for a score of 51 out of 150 is 6.48. meaning there is a less than 2% chance of getting that many answers right by chance.
I was so shocked that I tried Japanese, which I only studied for five lessons in 1996 before dropping it. I only got 18 points. That gives me a p score of 13.52, and a probability of 0.0025. In short, I am so bad at Japanese that there is a less than one in four hundred chance of getting so few right by accident. I realised, after completing the test, that I had consistently chosen the answers that contained a Japanese word that I thought I had heard before, and perhaps by doing so I had actually managed to do worse than random chance.
Of course, it's very possible that I've screwed up the calculations. I haven't done a Chi-squared look-up by hand in a long time. Still, I am forced to contend with the notion that I might actually have some knowledge of Irish without any conscious access to it - enough knowledge to be measurable. That's scary.
Update: The scoring isn't structured the way I thought it was. That may contribute to my odd score distribution.
Further Update: Yeah, I managed to confirm that I did screw up the math. There are 50 questions, not 150. That's a lot fewer independent variables. There is a 15% chance that my Irish score is pure luck if the points are distributed evenly over the questions and all my probability estimates are way, way off. So, there's an 85% chance that I actually know some Irish.
Thursday, August 07, 2003
Well, it looks like yet another actor will be running for office. I'm sure everyone in the English speaking world and many who aren't now knows that Arnold Schwarzenegger is running for governor of California. Unfortunately, Gray Davis is not really responsible for California's running disasters. He didn't create the energy crisis, the dot-bomb crash, or Prop. 13. Whoever wins inherits this mess, and won't have anymore power to fix it than Davis does.
You know, I remember my undergrad PoliSci prof complaining that liberals think there's some kind of button on the President's desk that just makes everything better, and for some evil reason he's just refusing to push it. Well, what goes around comes around. Does anyone really think that there's a button on Gray Davis' desk that just fixes things? Does anyone really believe this election changes anything? The legislature remains Democrat, and California is too big and too powerful for the state reps to just roll over and play dead for an actor in the Governor's Mansion.
Wednesday, August 06, 2003
And in weather, it's hot and dry...
...and by all evidence likely to stay that way.
Since European penis stories seem to garner more hits than anything else I write, I thought I might continue in the same vein with this from the BBC.
Skirting around Sweden's heat
Now folks, Umeå is just south of the 64th parallel north. That's just a bit further south than Fairbanks, Alaska or about as far north as my childhood home in Iqaluit. If it's hot enough that far north to make a bus driver wear a skirt, think of what it's like in semi-tropical climes like London and Amsterdam, or what I have to put up with. It seems that Belgium has a law that anyone whose workplace is over 30C can go home with pay. Unfortunately, my office is air conditioned.
Monday, August 04, 2003
In search of the standard Euroschlong
Ananova is carrying this rather juicy bit of news from the annals of European standardisation:
German penises 'too small for EU condoms'
That certain European states might, umm, select measurement methodologies with higher than industry standard margins of error is understandable. I mean, it's hard to get honest information about anything related to sex - a well known problem in the social sciences. Besides, I have to ask myself how the measurements were taken. I mean, if I had to get it up for some German urologist with a caliper, I imagine I might well be a bit subpar in the fully aroused department. If one wanted to get answers for the fully erect size of the German wiener, I suppose getting busty young nurses to take the measurements might raise the numbers a mite.
But what I want to know is where in hell this 18 litres number comes from? I mean, Jeebus! How big do Euronads have to be before you get anywhere close to that safety margin?
Kjenn Jie Noch Plautdietsch?
For those who have an interest in languages and who have been reading my posts from my Grandfather's papers, a guy in my office with a passion for Lower Saxon passed me a link to the online version of the only Mennonite Plautdietsch dictionary I know of. I have a copy of the paper version of Kjenn Jie Noch Plautdietsch? and the author/editor Hermann Rempel is, I believe, one of my many, many second cousins on my mother's side.
People have spelled Plautdietsch very differently over the years and there is still no fully agreed upon standard spelling scheme. Like most Mennonites - who in the old days used German as a liturgical language - the writing scheme is based on standard German, although the sounds are much more Dutch-like.
Looking through the word lists gives me warm fuzzies. When I was a child, my mother used to say Schlop scheen! to me and my brother every night before we went to bed. I still have a taste for wreninkje, but I doubt I'll find any in Belgium.
Islamic Finance in America
I have something of an amateur interest in Islamic finance. The main reason is an eclectic interest in social orders that differ sharply from Anglo-American liberal capitalism. This isn't because I think there's anything uniquely evil about "Anglo-American liberal capitalism" - quite the contrary - I just don't think it's the final, natural, perfect form of human society.
An AP article on Islamic mortgages in the US just came over the wires, and I thought it might be interesting to my readers:
Muslims Embrace New Finance Alternatives
The way an Islamic mortgage works is that the bank buys your house and sells you a share of it equal to your downpayment. You pay rent on the house at a rate fixed in your lease as well as a little bit extra that goes towards purchasing the house itself. If you have kept up your payments, at the end of the term you own the house outright. In principle, this is little different from a fixed rate mortgage. The main virtue that it has to offer, I think, is that it is a little more honest. People with mortgages fool themselves into believing that they are homeowners, and an Islamic mortgage eliminates that illusion.
The main principle of Islamic finance is the interdiction on interest. This means, as I understand it, that banks can charge for whatever services they do for you, but they can not make those fees proportionate to the amount of money involved in the transaction. They can, however, take business risks and reap proportional rewards.
This means that taking out a business loan from an Islamic bank makes the bank a limited partner in your business and therefore able to collect a portion of your profits, although it also means it has to swallow your loses. This actually makes a good deal of sense to me and it has some resonance with the way I understand commerical finance to work in Japan. One of the factors in the establishment of the Japanese keiretsu structure was the lack of reliable credit rating institutions. Becoming part of a keiretsu means that you have a special relationship with a bank and often a "trading company" - which is a sort of middle man who acts to ensure contracts to supply goods and services. This relationship can act as a sort of guarantee of your creditworthiness. The structure of Islamic commercial banking actually seems to fit this model pretty well, and it has worked very well for Japan, Korea and Taiwan. Japan's recent troubles seem to have more to do with the implicit guarantee that no company will ever be liquidated - no matter what the cost - rather than the actual structure of the Japanese firm.
Business loans for capital purchases work on the same sort of principle as Islamic mortgages. The bank buys the equipment you want and rents it out to you on a rent-to-own sort of plan. This too makes matters a little more honest. I do wonder if Enron would have been able to so easily hide its true financial situation had it been compelled to declare all the things it borrowed money to buy as leased assets and to give each of its creditors a share of the company's profits and full access to their books.
Now, I am not by any means claiming that interest is inherently evil. I am not a Muslim and I have little interest in whether something is forbidden by the Quran. There is a wonderful post on D2 Digest on the importance and usefulness of interest. I am, however, very interested in how alternative finance mechanisms work and how well they serve economic needs. One of the points D2 raises is the importance of interest-bearing loans in personal finance, and especially how important it is when you don't have assets:
Loan sharks seek out poor neighbourhoods; they don't create them, and the fact that extremely poor people are nevertheless prepared to pay extortionate prices for the ability to move consumption around in time just confirms what a great thing it is to have access to debt.
Islamic finance seems to be able to build homes and businesses. It looks to me like it is no more inherently inefficient or problematic than traditional finance for these purposes, and may have the virtue of keeping businesses more honest. However, personal finance - the most oppressive aspect of banking in terms of its potential negative impact on people's quality of life - still seems to be a real problem for Islamic banks.
The Abu Dhabi Islamic Bank is something of a trend-setter in the Islamic bankng world, and does offer some personal finance products. (They also offer some strangely anachronistic services, like Ladies' Banking, whatever that is.) Their credit card offerings look a lot like American Express: at the end of the month, you must make full payment. However, take a look at their "Al Khair Financing" product. I'm not an Islamic theologian, but I can spot when a loophole violates the spirit of the law as well as the next guy. Unless there is a clause here that I'm not seeing, this looks a whole lot like a fixed-rate loan to me, except that the bank has transfered a portfolio of commodities to the borrower instead of cash money and calls the interest payments a "fixed profit rate." There appears to be no real risk sharing at all, just avoidance of "interest" by the narrowest possible definition.
Anyway, whatever virtues Islamic finance may have in commercial banking and loans for the purchase of secured assets, I haven't seen any way for individuals to get unsecured credit in a genuinely interest free fashion. The only mechanism I can think of that would fit the bill is for the bank to give you a loan in return for a percentage of your personal income for a fixed length of time - something like the PAYE system for student loans in Australia and New Zealand. That way, the bank's loan to you is essentially an investment secured by your earning potential, and the bank bears the risk that your earnings will not be adequate to make the payments.
The thing is, I can't really imagine people agreeing to it. It's too invasive. One advantage of interest is how impersonal it is. The bank cares very little how you lead your life, as long as they get paid. On the other hand, if the bank has a vested interest in your employment situation and you become unemployed, you have a very powerful advocate with every interest in following up on wrongful dismissal charges. Furthermore, the bank's bottom line is quite directly linked to the general employment situation in your industry and area, and they have a vested interest in making preferential loans to any business that might improve the local employment situation. Perhaps the personal security that follows from such a relationship has advantages to outweight its intimacy.
Anyway, just some thoughts while I work through my writer's block on the stuff I'm being paid to do.
Sorry to have been out all weekend. I ought to write a response in the comments to the post below and I have an e-mail about Activity Theory from a correspondent in Rochester that I need to write back to. In addition, I promised a part two to the post below.
None of these things happened this weekend and they aren't going to happen today. I have had a relapse of my Belgian summer allergy problem, in addition to acute Belgian bureaucracy issues, compounded by a writing assignment at work for which I am feeling absolutely zero inspiration. So instead, let me point to an ugly trend that is making me hesitant to travel in the US again.
Tightening of Visa Rules to Disrupt Travel for Some
Not every country in the world allows visa-free transit, however a great many do. Such a policy is a real boon to a country's airlines, and US companies have profitted heavily from free transit policies in New York, Miami and Los Angles, primarily for travel to and from Latin America. If this is going to be a "pro-business" administration, it would nice if they were at least consistently pro-business.
US anti-war activists hit by secret airport ban
This is a bit worrysome for me. Although I have no particular associations with any political groups at the present time, I have noticed a lot of bots with *.gov and *.mil addresses hitting this blog. I have no evidence that opinions voiced on the 'Net are being used to build up dossiers, but I worry that they might. It makes me somewhat scared to travel the US.
What I'm most scared of, since I'm not a US citizen but I do have a resident's permit, is this:
Considering how little the Canadian government has managed to do for its citizens in the past when the INS gets ahold of them, I have to question whether my nationality will do me much good in the US anymore. At least I can't be sent for interrogation in Jordan and then deported to Syria, at least, I don't think so.