Sunday, May 22, 2005

Salvation Army Book Fair

To all readers from Singapore...

You mayn't know it, but the Salvation Army has a pretty big stock of books that people have donated. These books are available for sale at various branches, the biggest (for books) being the Upper Serangoon store. The book section is entirely run by volunteers from the Salvation Army Library Enterprise, and we will soon be holding a big sale (yes, it's the SALE sale) during the June holidays to sell off our excess stock. Details as follows:

Date: 3-4 June (Fri & Sat)
Time: 11 am - 7 pm
Venue: Geylang Methodist Secondary School Hall
Getting there: Map
MRT: Aljunied
Bus: 21, 26, 40, 51, 62, 63, 64, 67, 80, 100, 125, 155, 158.

We have books for children, young adults, adult fiction, non-fiction and reference - you name it, we have it. Some books are almost brand new and can be found selling for 20 times the price at bookstores; others will be much older, out-of-print and impossible to find anywhere else. And we will have lots of books at the sale: possibly up to 20,000 books (we haven't counted exactly) rolling out over the 2 days. There will also be children's activities from time to time such as lucky draws and storytelling.

The books will be dirt cheap: all of them go for $1, or 50 cents (if marked 2 for $1).* If you come across a bunch of small books and don't think they're individually worth $1, you can also approach a volunteer to have them marked down. Of course, all proceeds will go to other projects run by the Salvation Army.

Please come! Because the whole thing is being run on a shoestring budget, we haven't the money for big publicity - so please spread the word to your book-loving friends too. Also, because this is entirely volunteer-run, please forgive us if the sale isn't as professional as what you may be accustomed to. We hope the low prices and wide range will make up for any difficulties you may encounter during the sale itself. We also welcome help from anyone who is free on the days of the sale, or even interested in volunteering at our Saturday sessions at Upper Serangoon after the sale. Drop me an e-mail if you're interested and I'll put you in touch with the right people.

Link: PDF brochure for the book fair

*This is the current pricing policy; things may change before the day itself (and will be duly noted here).

UPDATE: Well, we managed to get $3000 out of the book fair, apparently. Not bad for a few mornings' work! Thanks to everyone who came and contributed!

Thursday, May 19, 2005

Time scale modification - the $0 version

A recap of yesterday's post: You can speed up speech (and music) without making people sound like chipmunks, and slow them down without making them sound like Darth Vader. This is called time scale modification. Apparently playing things faster is good for you because (1) it saves time and (2) it forces you to concentrate, so you learn better. But playing things slower can have a benefit too, when you're listening to, say, radio broadcasts in a foreign language that you're trying to learn. Often these go too fast for you to catch much, but slowing them down to 0.7-0.8xRT makes them perfect. One way to do time scale modification is to use Enounce, a plug-in for Windows Media Player and RealPlayer that lets you play stuff anywhere from 0.3-2.5xRT.

Today I'm going to look at another way of doing TSM, as it's commonly abbreviated. It's really quite easy. First, head over to and download yourself a free copy of Audacity. Audacity is open-source software for recording and editing sounds and works in Windows, Linux, OS X, you name it. It's really nifty for manipulating audio - for example, you can mash recordings up by highlighting an area of the recording and then dragging it off somewhere else. There are other free sound editors out there, too: Audacity is just the one I have installed on my computer.

Now, how to do TSM in Audacity. Select the whole recording, then go to Effect, and Change Tempo. Don't use Change Speed, that gives you chipmunks and Darth Vaders. Change Tempo shortens (or lengthens) your audio without changing the pitch - just what Enounce does. Just move the slider to where you want the speed, and press OK. It takes a few seconds to work its magic - though you can preview the effect first just to be sure that's the speed you want - and press play. Et voila! Audio speeded up or slowed down as you wish.

I don't think that the TSM effect is quite as smooth as in Enounce, which is only to be expected - this is free software, after all. There are other drawbacks, too: you can edit Ogg Vorbis, MP3 and WAV files, but nothing else - so if you have a file in .wmv format, or RealPlayer format, the only thing I can think of for you to do is to play the file and record it simultaneously in Audacity. Which pretty much defeats the purpose of the exercise in the first place. And many Internet stations encode their files in those formats. Of course, for language-learning purposes it's not so bad to listen to anything twice. Practice, practice, practice!

So I might just overcome my habitual skinflintedness and fork out $29.95 for Enounce, unless anyone knows of a way to open up files of other - proprietary - formats in Audacity or any of the other free audio editing programs, like Praat or Wavesurfer. Now that I'm running my free 7-day trial version, I don't think I could go back to listening to audio at regular speeds again. It would just be too boring.

Wednesday, May 18, 2005

Time scale modification for language learning purposes - the $29.95 version

I don't know about you, but I'm a really impatient person. I read fast because I don't like to spend too much time reading. Now, I would really like to be able to listen to more lectures and interviews and other audio online, because there's really good stuff out there, and coming from a college town back to Singapore, where there's not much in the way of public lectures and all that sort of thing (well, none that I'm interested in hearing, anyway), it's the next best thing.

The trouble is, I'm too impatient to sit down and listen to it. When it's not live, when you can't see the person who's talking, you realise that people actually speak really slowly. They are careful to enunciate their words - all well and good, but I've a pretty good language model in my head, I can stand some speeding up.

Now, supposing you were listening to a lecture and pressed the fast forward button - what do you hear? Chipmunks? Mickey Mouse? Well, something along those lines anyway. That's because playing it twice as fast makes the frequency twice as much, and frequency is correlated with pitch. Playing something twice as fast makes the speaker's pitch rise - hence the "chipmunk effect". I'm not kidding, that's what people call it. Now, to speed things up *without* creating the chipmunk effect - you'd probably need to take each vowel and truncate it, etc., etc. - I haven't looked into how it works, but there are tools out there to do this, and it's called time scale modification.

So I looked on the Web and found that a company called Enounce makes tools to speed up your listening. The software is a plug-in to RealPlayer and Windows Media Player, and works whether you're playing something you've downloaded or streaming from the Net. Although, of course, if it's live you can't play it twice as fast - it would be really funny if you could, but you can't. There's a free 7-day demo, so I tried it out on a talk from IT Conversations by Cory Doctorow. It worked pretty well! No squeaky voices (though you can alter the properties to get the chipmunk effect, if you need a laugh). I found 2x was a bit fast for me, I wasn't used to it yet. 1.5x was perfect this time round. If I practise with it I can probably work up to 2x or higher, I guess.

Apparently there are other benefits besides the time factor to this software: students learn better because playing it faster makes you concentrate harder, for example. [Link to BYU paper on this, pdf format] But what interested me more was seeing that you could not only speed up, you could also slow down - to as low as 0.3x the regular speed.

Many language learning advice-dispensing websites advise that you should try listening to online broadcasts in your target language. These are available for an incredible array of languages. Deutsche Welle has 30 languages, the BBC World Service broadcasts in 43; Radio Canada 8. (These are my favourite.) Each of these three has broadcasts in Arabic, my target language - the only thing is, the presenters speak too fast for me. I'm not yet at the stage where I can listen to broadcast news read at normal speed, though when I know the story in question I can usually work it out. What I really need is for them to go just a little bit slower. The BBC knows this, and Deutsche Welle knows this - both have programmes for language learners where they speak a little more slowly, and clearly, using less difficult words. But I haven't found anything comparable for Arabic.

But all I have to do is to put Enounce and those broadcasts together, which I tried today. I found that playing at 0.7 - 0.8x the normal speed worked pretty well for me. I had time to work out how the more complex words parse, and to recall what they meant. I think this will be a big help for my Arabic-learning efforts.

The thing is, Enounce costs $29.95. I think this is a pretty reasonable price, but tomorrow let's see if we can go one step better and do this for free.

Monday, May 16, 2005

NLB xISBN bookmarklet

I forgot to post this for the longest time, but here it is at last:

There's now an official OCLC xISBN bookmarklet for the National Library Board, Singapore! You can install it here. Basically, Eric Childress of the OCLC noticed my blogpost about modifying my NLB LibraryLookup bookmarklet to take advantage of the xISBN functionality, which lets you search for various editions of a book, which have different ISBNs, with a single ISBN search. That is to say, my post was about my inability to modify the bookmarklet. Jeff Young, also of the OCLC, then created a bookmarklet specially for the NLB catalogue and e-mailed me about it. Thanks, guys.

Sunday, May 15, 2005

Nature observations

I've been engaged in some naturalistic pursuits lately. One, a rather curious caterpillar landed on the family car the other day and we took it and put it in a jar. Everything about it seemed quite ordinary except that it had a big bulbous green head. It looked almost as if it was a caterpillar trying to consume a pea, but it was quite clear that the green head was fused to the rest of its body. By the next morning, it had not eaten any of the leaves we put in for it, but had produced copious amounts of waste, to put it mildly. About 10 pellets. And by the afternoon, it had become a yellow cocoon! Unfortunately, I was out the whole day so I didn't see it weaving itself in. I wonder how long they stay in their cocoons, and what sort of butterfly it will become. If any readers recognise my description of the caterpillar, let me know!

The second experiment is an ant colony. I bought, rather on impulse, this Antworks set a few weeks ago. (Read the linked website for a full description & pictures.) We set out on an ant-collecting expedition today, and quickly discovered why Antworks offers packets of ants for sale. They're devilishly quick! It took a while to gather just 11, and that's nowhere near the 20-25 recommended. The ants we got also seem a bit on the bloodthirsty side. There are ant parts lying around the colony, and at least two of the ants seem to have been badly damaged. On the plus side, though, they seem avid diggers, so hopefully we'll see some beautiful tunnels - if they don't kill each other first. Another advantage is that they don't seem to have climbing feet, and can't climb up the walls of the colony, which makes it a lot easier to put new ants in, unlikely the ants we find in our kitchen. From my observations today, it looks like they have a cooperative system, with one ant digging down below and other ants passing the debris up. Very Great Escape-ish.

Sunday, May 08, 2005

Game theory

The BBC Magazine has an interesting article on how to win at everyday games like Connect Four and Reversi. One surprising one for which I would have thought no strategy possible is Scissors, Paper, Stone (the Americans call this Rock, Paper, Scissors as I recall). This was the game that decided which of the two great auction houses, Sotheby's or Christie's, was to have the opportunity of auctioning off Maspro Denkoh's $20 million collection of Picassos and van Goghs. Christie's consulted the 11-year-old daughters of one of their directors, who proposed the following strategy:

1. Stone is the one that "feels" the strongest
2. Therefore a novice will expect their opponent to go for stone, and will go for paper to beat stone
3. Therefore go for scissors first

I think my initial choice is rather random, but from now on I may keep their advice in mind.

Another interesting article I've read over the past few days is "On the Survival of Rats in the Slush Pile" by Michael Allen (also known as the Grumpy Old Bookman). It's a little monograph - almost a screed - on the publishing industry and what's wrong with it. Basically, the methods publishers use to determine the value of a book are completely wrong, and the machinery behind publishing is badly broken, and if you want to become a writer and earn fame and fortune, like Dan Brown or J.K. Rowling, you might as well give up now. He also suggests methods for reforming the industry, and his basic advice to would-be writers is: go ahead and write, but don't give up the day job.

It's a very quick, informative read. I especially liked how he threw in some insights from statistics, in particular Nassim Taleb's books about chance as well as the theory of the Long Tail.

I wish more books were like that - short and to the point. What happened to the old practice of writing monographs rather than books? I see so many books in the bookshops these days that talk about interesting subjects, but - honestly, who wants to read 200+ pages about, say, Daylight Savings Time? Or salt? Can't it be compressed down a bit? And made cheaper? Then I could spend less time finding out more, with less pressure on my wallet.

Writers wouldn't have to spend so much time fluffing out their books, either - they could just state their piece and be done with it, sell it for less but probably have more readers to make up for it. We need to make <100-page books fashionable again. They would be perfect to distribute electronically, too. "On the survival..." is 70 pages long, and doesn't hurt the eyes. I couldn't imagine reading 200+ pages at a stretch at the computer. But 70 pages, with generous margins, is pretty good, with a few minutes' break at the halfway mark.

Friday, May 06, 2005

Reading Chinese characters

Here's an interesting article about the psychology behind learning to read. Especially interesting are the experiments it describes about halfway down about reading Chinese, which are taken from Spinks et al., "Reading Chinese characters for meaning: the role of phonological information".

Many people believe that the Chinese script is ideographic. [See John DeFrancis' The Ideographic Myth for a refutation.] But some neat psycholinguistic experiments demonstrate that Chinese characters really represent sound as well as meaning, as described in the ScienceNews article.

[In the following, I describe the different cases they examined as different experiments, but really all the possible cases were presented together in one experimental session. I call them 'experiments' for my own convenience.]

Experiment 1: the Stroop effect. If we are shown a word, e.g. RED, and are asked to say what the colour the word the printed in is, it takes us a lot longer than if the the word and the colour matched, as in RED. You can experiment with it at this site.

Turns out that Chinese has the same effect. Reading the word (meaning "red") takes Chinese readers longer than if the colours matched. This establishes the baseline effect.

Experiment 2: If we give a word in Chinese that has the same pronunciation as , hong2 (that's the syllable "hong" in the second tone), such as (meaning "flood"), in an incongruous colour, a similar delay occurs! One of the experimenters, Charles A. Perfetti, of the University of Pittsburgh, explains:

"This effect indicates that written characters correspond to sounds in spoken Chinese, not to specific words. The pronunciation of flood calls to mind red and slows naming of the clashing ink color, Perfetti says. If the characters represented specific words, instead of sounds, this delay would not occur."

Experiment 3: If we give a word in Chinese that is of the same syllable but different tone, such as ("boom", hong1 - again the syllable "hong", but in the first tone), then there is still a delay, though it is now smaller. So evidently reading this word calls up the pronunciation "hong", which recalls the similar-sounding 红 and its meaning, red, which inhibits the correct naming of the colour the word is printed in.

Experiment 4: This one used only semantic information, providing a word that is very closely associated with the colour. For example, for the colour red, they provided the word for blood, (xue4). For blue, it was 'sky', . This had about the same average response time as in the second case, where a word with the same pronunciation but totally different meaning was supplied.

Experiment 5: This worked in the other direction, testing whether giving a word that sounded the same as the colour it was presented in helped. For example, they provided the word ("ponder", ) which sounds the same as the word 'green', 绿,also pronounced lü4. This shortened the response time as compared to a random word presented in the colour green, though not to the extent that it was the same as presenting the colour name itself. So semantic associations could be triggered by phonological similarities. Actually, I'm not so sure about how this experiment helps so much - it seems to me that the effect can also be explained by the fact that there isn't another totally different pronunciation to interfere with identifying the pronunciation of the colour.

But anyway, to sum up, the findings are that:
(1) phonological information is obligatorily activated when reading Chinese characters, even when it is not in the reader's interest to do so. At the very least, the pathway is parallel between triggering of phonological information and triggering of semantic information (this is called the "parallel activation hypothesis").
(2) Phonetic similarities trigger semantic associations, but not to the extent that complete orthographic identity does.

I think it's a neat effect. It's a very clever use of the amount of homophony in Chinese characters.

Also interesting is the following paragraph, from the ScienceNews article:

Studies of blood flow and electric responses indicate that Chinese readers activate many of the same left brain areas that English readers do, Perfetti adds. Right brain regions involved in vision also contribute to reading Chinese but not to reading English. This finding is consistent with the possibility that learning to read Chinese stimulates spatial perception.

I wonder if Chinese speakers are better at finding their way around, then?

Baby on board - sometimes

Is this lying? I saw a car today with a little sign saying "Children on board". But there weren't any children actually in the car at the time. Now, the little sign seems to be making a pretty bald statement that two or more children are in the car. Obviously, in real life you don't take the little sign down every single time the children leave the car, and put it back up when they get back in. So what the sign really means is "Children on board some of the time", even though I'm thinking most people when they read it will interpret it to mean that there are. So are these people lying? (I suppose you could say the sign is lying.) Or are our expectations managed by the static-ness of the sign?

Monday, May 02, 2005

Perceptions of randomness: survey

Haven't had a whole lot of linguistic insights recently, so posting has been light. I've been working on writing up some notes for learners of Arabic. The notes are mostly for intermediate to advanced students. I'm writing up some on the verbal morphology right now. I may post them online when I'm done with them.

And now, to the main topic of this post. I've created a very short survey for testing people's perceptions of randomness. It consists of just one question. Take a look at it, and please fill it in. I'm using SurveyMonkey, and it's a free account, so I can only collect 100 responses. If (& when) 100 responses are in, I'll post the results. (I'm not sure if answering the survey brings you to a page where all the responses to date are displayed. I don't think so. Let me know if I'm wrong.)

Oh yes, one warning: the survey is image-heavy, and may take a while to load.

Click here to go to the survey. If you want to know more about the rationale behind the survey (and the correct answer), click here.