Tuesday, June 28, 2005

"The Sparrow" by Mary Doria Russell

I'm currently reading "The Sparrow" by Mary Doria Russell, which is science fiction but/and includes some linguistics stuff. I was inspired to by this post by Suzette Haden Elgin, herself a linguistics/science fiction writer. The book has been amazing so far (I'm halfway through it). She writes wonderfully, and it's all very interesting, and above all suspenseful.

What you know right from the beginning is that the Jesuits sent a team of priests and "civilians" to make first contact with a planet called Rakhat, something went terribly wrong, and only one man returned, physically and emotionally scarred. But why, and how? Even as you read about the inquest into what happened, a second storyline starts tunnelling its way from the beginning of the story, telling you all about the people who went on the planet and how the expedition got started. And the painful thing is that you know that every one but one is going to die. But you've started to like them, and it hurts.

Anyway, there's quite a bit on languages in it so far. I liked this bit, where the protagonist, Emilio Sandoz, describes one of his language-learning techniques:

..."Sometimes," he told her, learning forward over the table, speaking without realizing how it would sound, "I begin with songs. They provide a sort of skeleton grammar for me to flesh out. Songs of longing for future tense, songs of regret for past tense, songs of love for the present."

He blushed when he heard what he'd said, making it worse, but she took no offense; indeed, she seemed to miss any connection that might have been taken wrongly. Instead, she seemed struck by a coincidence and looked out the cafe window, her mouth open slightly. "Isn't that interesting," she said, as though nothing else he'd told her so far had been, and continued thoughtfully, "I do the same thing. Have you noticed that lullabies nearly always use a lot of command form?"...
This reminded me of a song we learnt in Arabic class: حبّيتك و بحبّك و هحبّك على طول was the first line, meaning "I loved you, I love you and I'll love you forever". It was kind of neat for remembering the distinctions between the three tenses in Egyptian Colloquial.

Sunday, June 19, 2005


Well, I'm back from a nice holiday in the States, which explains the lull in postings. While there, I discovered Yubnub, which bills itself as "a (social) command line for the web". Basically what happens is that you go to the website, type in a command like (for example) "wikipedia callosum" and get the Wikipedia search page for the word "callosum". Cool, right? What makes it cooler is that you can transform your Firefox search bar into a command line, so you can type "wikipedia callosum" straight into the URL bar and get the same result! Coolest of all is that you can submit your own commands, which is what I've been playing around with for the last ten minutes.

Here are the commands I've created:

Translates a word from English to Arabic using the open source Qamoose dictionary.
Example: eng2ar anything
will give you the translation for the word 'anything'.

Searches for books in the Online Books Page (http://onlinebooks.library.upenn.edu) by words in the title. If you wish to search by author, use the Yubnub command freebook-author.

Example: freebook beagle
Example: freebook voyage beagle

These searches return "The Voyage of the Beagle" by Charles Darwin.

Searches the Online Books Page (http://onlinebooks.library.upenn.edu) by author name.

Example: freebook-author darwin
Example: freebook-author charles darwin

Searches the National Library of Singapore catalogue by keyword.

Example: nlb harry potter

I also tried (and failed) to create one for the National Library catalogue, nlb, but it does NOT work as advertised. It can only take in one search term at a time, and not two (like "harry potter"). nlb now works as advertised - thanks Jon! You can type in "nlb harry potter" to do a keyword search on the National Library (of Singapore)'s catalogue. (If you want to know why it didn't work before, scroll down to the bottom.)

Other EXTREMELY COOL commands (that I've come across so far) are:

g - searches Google.
gs - searches Google Scholar.
shrink - TinyURL, but I'll probably stick to my regular bookmarklet most of the time for this. You can find it on the TinyURL main page if you scroll down a little.
gmaps - Google Maps.
map - Multimap.
am - searches Amazon.
tr - Google Translate. Use it like this: tr en fr friend to translate 'friend' from English to French. Other languages: simplified Chinese zh-CN, German de, Italian it, Japanese ja, Korean ko, Portuguese pt, Spanish es.
thes - thesaurus.com search
say - text-to-speech. (Cor blimey!)
acro - look up the definition of an acronym.
answer - Answers.com

Three cheers for Yubnub. I think I'll be using it a lot.

nlb doesn't work:
Yubnub automatically inserts a + between two or more search terms. So "nlb harry potter" gets the URL <> which is incorrect. You want "_harry%20potter" at the end instead. The search will still return results, but it will think it's searching for the word "harry" only.

I've no idea how to fix this, but I emailed the site's creator, Jonathan Aquino, about it, to see if he can fix this. I suppose you could try using the catalogue available through elibraryhub, maybe that would work.

UPDATE: nlb does work. Jon specially implemented a new syntax that transforms spaces into, well, spaces (%20) rather than into plus signs. And the NLB catalogue is immortalised on the yubnub website as the exemplar for this syntax. Jon was also kind enough to credit me for "suggesting" this fix, but it was his idea, I just supplied the problem!