Sunday, September 20, 2009

Best language-learning environment for kids: Two parents chattering away with fancy words

An article in the New York Times entitled Birth Order: Fun to Debate, but How Important? discusses language learning and birth order. Here's the money quote:
Frank J. Sulloway, a visiting scholar at the University of California, Berkeley, and the author of “Born to Rebel: Birth Order, Family Dynamics, and Creative Lives(Pantheon, 1996), points out that second-born children tend to be exposed to less language than eldest children. “The best environment to grow up in is basically two parents who are chattering away at you with fancy words,” Dr. Sulloway said.
That seems fairly obvious to me—expose them to more words and they learn more words—but the good doctor lends some authority to it.

Labels:

Tuesday, September 8, 2009

Get Chinese pinyin for any text

While I couldn't find the equivalent of Hiragana.jp for Chinese (i.e., letting you convert an entire website and then browse it), there are a few good converters that will take a block of text and convert them from Chinese characters to pinyin for you.

There are lots of them out there that simply do it character by character, but as context can affect the tones (e.g., whether the characters are part of a word or not), the best ones take this into consideration.

Here are a two of the better ones:

Labels: , , , ,

Get Japanese furigana for any website or text

If you're looking to add furigana to an entire Japanese website, just drop the URL into Hiragana.jp. It'll add furigana to the entire page for you, and then you can click through the website normally and have furigana on every page.

If you've got a block of text that isn't on a website for which you want the furigana, then simply copy and paste it into Furiganizer. It'll do the same thing that Hiragana.jp does for websites for whatever block of text you dump in.

Labels: , ,

Sunday, September 6, 2009

44 Chinese characters you may need to relearn

Great... just when you think you've checked off all your Chinese character boxes, China starts debating whether to change the boxes by modifying certain characters.

Read more... For the first time in some two decades (in which time, I'm sure, many of us have commenced our studies of Chinese), China is considering revising some characters. The revisions only affect 44 characters, and below you'll find them in their revised format (click to enlarge):



Some of these characters I doubt whether I've even seen before, while a bunch are quite familiar. At first glance, I didn't even notice the changes, but, on a second look, it seems that one of the biggest changes is that of 朩 děng becoming 木 . So clearly we're not dealing with a revolution here, but still something that we learners should take note of.

And all this just when Taiwan starts talking about making the jump to simplified.

I don't know about them, but I'm still holding out for the implementation of one of the few ideas from Mao Zedong that just might make sense:
Reform of Chinese characters must be carried out in the direction of total Romanization that I believe is the mainstream of language development in the world.
Then again, he also thought that backyard steel furnaces would be a good idea, and that didn't turn out so well, so maybe we should just stick with the characters and consider this idea to be in the official 30% of things that Mao got wrong.

Links:
通用规范汉字表 Tōngyòng Guīfàn Hànzì Biǎo (General-Use Standard Chinese Character Table) (in Chinese) [中国语言文字网 Zhōngguó Yǔyán Wénzì Wǎng (China Language and Writing)]
Revision of 44 Chinese Characters in Hot Debates [CRI English]

Labels: ,

Get the text of Japanese podcasts with Podcastle.jp

The Japanese government seems to be doing a few things that are pretty useful for language learners. I noted a few while back that they've made the official Japanese-English dictionary of legal terms. That unfortunately is probably only of use to lawyers and the like, but last week I discovered another gem, again courtesy of Japanese tax dollars*, that is of more general use for Japanese learners: Podcastle.

I've been trying to find some Japanese podcasts for which the text is also available, without much success. LingQ's list of resources surprisingly has nothing suitable, and googling was turning up little. For whatever reason, there seem to be few Japanese podcasts that also provide transcripts.

But then I stumbled upon Podcastle.

Read more... Here's how Podcastle.jp describes itself:
Podcastle is a service that lets you search the audio of podcasts in Japanese. Voice recognition technology converts the audio into text. Users can then freely edit any recognition errors.
So, basically, they use a less-than-perfect voice recognition technology (because, after all, I don't think one exists yet that gets close to 100% accuracy) and then users edit the computer-generated transcripts to fix errors.

As I've noted before, one of the Japanese-language podcasts I've been listening to regularly is Yoichi Ito's Business Trends. And, sure enough, it's on Podcastle.jp. I went through a few of the transcripts, and overall their accuracy is pretty good. Indeed, many of them have hundreds of corrections. That tells me that the crowdsourcing is working well, but also that the voice recognition technology must leave quite a bit to be desired.

However, even with the crowdsourcing, the transcripts are not completely accurate. I was listening to one podcast and I noticed that a term appears to have been outright skipped. The term in question was スローダウン suro-daun, which means "slowdown" in the sense of the economy slowing down and is taken directly from the English term. The term was said quickly and somewhat quietly, and I could see why the voice recognition technology might have missed it, but I had no trouble understanding it so I'm sure native-Japanese speakers are able to hear it just as easily—but it remained completely omitted from the text. (I became a crowdsourcee by making the addition myself.)

Another cool feature of Podcast.jp is the ability to follow along with the podcast. You can play the podcast and Podcast.jp will indicate what text you are currently listening to. It's not completely accurate, but it's usually within a few words of where the audio is.

Until something more accurate pops up, this is a pretty good way for Japanese learners to get audio combined with text. The biggest problem is, of course, that the text doesn't always match up with the audio, so it's helpful if you know enough Japanese to figure out when the text might be screwed up. But by and large, it's accurate and good enough to help you get many of those terms you didn't quite catch in the audio.

Despite it's limitations, I'd love to know about any similar things that are available in other languages, so if you've got the info, please drop a line in the comments below.

P.S. If you're wondering how I figured out that this is a government-run project, take a look at their "Credits" page (in Japanese):
Podcastle is released as the research results of the Podcastle Project of the National Institute of Advanced Industrial Science and Technology, an independent administrative agency.
So if this was made with taxpayer dollars, I wonder if there's a way to get the research results—and the code—for free. If there is, someone with skills in working the Japanese bureaucracy please do so, and make this available for all languages ASAP.

Labels: , , ,

Language-learning schools as a front for crime?

I'm generally a big proponent of using your target language as much as possible. However, I definitely strongly advise against using it to make fraudulent visa documents.

Link: Lawrenceville man sentenced for immigration fraud [Atlanta Journal-Constitution]

Labels: