Slashdot Mirror
Swahili Wiki-Dictionary?
Martin Benjamin writes "The Hartford Courant just published a feature article on the Kamusi Project Internet Living Swahili Dictionary. This project is using the Net to put together dictionaries that are as scholarly as any university publication, yet with a secure participatory model that draws on knowledge from users around the world. Now the project is developing learning tools that will build on the Kamusi model of collaborative scholarship."
How about a dictionary in Navajo or Iroquois? Heck, even pig latin would do!
Viable Slashdot alternatives: https://pipedot.org/ and http://soylentnews.org/
nope, you just have a chip on your shoulder. From the article: "With more than 80 million speakers in East and Central Africa, Swahili is the most widely spoken language in Africa, though a fully updated dictionary of the language has not been produced for 30 years."
"he vets every entry for accuracy, sometimes within minutes..."
How, exactly, does he do this? It sounds like quite a trick.
He mentions "Then there's the professional ecologist major in Benin - he's a birder. He's sent in hundreds of bird entries, every type of thrush or crow ever spotted in East Africa, with their English and Swahili names." How does he "vet" these entries if he's not an ecologist himself?
Wikipedia regularly receives all sorts of hoax and joke definitions, neologisms, fraternity-house in-jokes, and so forth. It takes more than "minutes" to sort some of them out.
Does he just go on his personal intuition, which entries sound right and "feel" right to him? Or what?
"How to Do Nothing," kids activities, back in print!
How is it racist? Be honest with yourself. How many people thing "high tech" when Africa comes to mind. Untill their ecconomy changes, I will still see the vast portion of Africa as tribal. Even as such, that does not make then any less human. So again, I fail to see how this story is racist. Sounds like your just walking on eggshells when it comes to political correctness crap.
Life is not for the lazy.
I seriously doubt that the computer:student ratio is better than that.
[Fuck Beta]
o0t!
I know a lot of college student who would use this. I for one have been using the yale kamusi project for a longtime. And hell yea african can use computers i know lots of them. Africa is not what you see on the discovery channel. When I came to this country I was appalled by the ignorance of American one of my teachers thought that Kenya was in the carribean and i had one kid ask me "how does it feel to wear clothes".
Just the other day, I was lamenting to my friend how the internet seems to have everything except for a good Swahili Wiki-Dictionary.
Looks like my chum went to great lengths to collect on our 50 Rand wager.
This can happen at Wiktionary (English version here). That is the first thing I thought of when I read the title of this articll the Wikipedia people thought of a multilingual wiki dictionary a while back, when thye still had to go around saying "please expand this article, Wikipedia is not a dictionary". I see that Wiktionary only has about 5 English entries for Swahili words. Hopefully this guy will make the content on his site available under a GFDL-compatible license so that it can be assimilated into Wiktionary.
"Benjamin compares his project to Wikipedia, the online encyclopedia drafted largely by a band of worldwide literati. He emphasizes, however, that, unlike Wikipedia, he vets every entry for accuracy, sometimes within minutes, before he posts them."
Yeah you could do that with Wikipedia too when it had 100 new entried a month, but once you reach 100 a second I'd like to see how he'd cope.
P.S. There is a Wiktionary in Swahili right here: http://sw.wiktionary.org/ It hasn't attracted too many contributors, what makes this guy think he can do better?
The dictionary does not re-create the language, it documents it. It is a "living" dictionary, meaning that it is designed to remain extremely current to the language as it is used, through the submissions of users who have their ears to the ground. However, only words that can be documented, through printed sources, radio broadcasts, contemporary Swahili music, etc, are accepted for inclusion in the dictionary. It is intended as a reference resource, not the word of God. As to whether anyone will know the difference about the accuracy of the entries, that surely depends on your definition of "anyone." The population of the Swahili-speaking world is roughly the same as that of the German-speaking world. Would you make such a comment about a project for German?
If you build it, they will come...
"We've done all the programming work that's possible, and I can envision hitting the print key in about two years," Benjamin said. -- Actually, that's a misquote. We've done almost all the programming work that's possible given our current budget (the project goes belly-up at the end of the year without further funding), but we've got a task-list/ wish-list a mile long. Why not find a few AFRICAN ORGANIZATIONS to pay for it? -- Simple - most African organizations don't have the money to fund this sort of work. Those that have the money invest in other priorities, like health and emergencies. If you know of any African organizations with funds to spare, by all means please let them know about the project! About Kamusi-in-a-Box: if this happens, it will be in association with the Tanzanian school system, and all the software would be going to schools that have already been set up with computers running the Swahili versions of Linux, OpenOffice, and Firefox. So yes, the market is there - the market is a whole bunch of computers at educational institutions around East Africa that are ready and waiting for learning content.
If you build it, they will come...
I seriously doubt that the computer:student ratio is better than that.
Not nessesaraly, and it doesn't need to be. Take for example Somalia, where *complete* deregulation (that is, no central government whatsoever) has lead to a telecomunications boom. The warlords may burn the books, but nobody burns the computers, because they are important to everybody.
So much so, that the BBC maintains a Somali language website: http://www.bbc.co.uk/somali
In fact, I am writing this message from (*gasp*) Africa! And how do you think all those internet scammers operate if they don't have Internet access?
If it's in you sig, it's in your post.
We are doing something similar for Chinese at http://www.adsotrans.com/ and it hasn't been a mistake opening the project to user contributions. When mistakes happen they generally tend to be because of human confusion in how the editing system works or the way the backend dictionary is integrated with other software.
So if the Swahili project is anything like ours, I'd assume the big issue is encouraging people to become active contributors rather than passive users. Their community of contributors is probably relatively small and generally self-selecting to people reasonably fluent in the language, so the system would probably be self-policing even without an elaborate software system governing access issues. The problems we face aren't technical issues so much as questions of finding the resources and time to improve the project.
So good luck for them in attracting funding/participants. And if anyone is studying Chinese please do check us out. We have a language-learning blog at http://www.newsinchinese.com/ which may also be useful to intermediate/advanced students looking to get away from their textbooks and savouring the poetic eloquence of the the Xinhua News Agency.
This was definitely lacking in the other initiatives.
Of course, this supposes that a committee of reliable people (typically, university researchers, professionals, etc.) culls the articles as they are submitted, and it does require a lot of time. They already do this for peer-reviewed scientific or technical journals, with the difference that they probably get paid for doing it.Still, I believe in a serious technical/scientific committee donating their time in order to review the validity of articles submitted to online encyclopediae, and being given the rights to prevent the modification of the online articles unless those modifications have been approved. This would be a great step towards reliability in the Wikipedia publishing process.
And besides, to compare this with another great cooperative project, would Linus Torvalds let pieces of the Linux code be updated by any anonymous coward without a proper code review done by a trusted person ? This is the direction that ought to be taken for Wikipedia.
In Soviet Russia, our new overlords are belong to all your base.
Nupedia, Wikipedia's predecessor, was exactly such a project.
You didn't hear very much about it because after two years and $250,000 invested, it had a grand total of "24 articles that completed its review process" and 74 more that were well along.
Many of Wikipedia's organizational principles and policies originated in Nupedia, and Larry Sanger maintains that the success of Wikipedia stemmed from the fact that it had its start in a community of people who were thoroughly steeped in Nupedia ways of doing things.
Still, it is hard to see how Nupedia can be described as other than a "failure."
"How to Do Nothing," kids activities, back in print!
Swahili is a lot less fragmented than kahei believes. "Standard" Swahili is quite widely spoken, and most of the terms in the Internet Living Swahili Dictionary currently are Standard. However, several other dialects (certainly not zillions) are spoken, and the project supports multiple dialects through its Edit Engine. At this point the Dialects feature is underused, but we are developing search tools to make the feature more useful and user friendly, so I'd expect increasing dialect information in the dictionary as the project goes forward.
If you build it, they will come...
pilgrim23, you are wrong. The history of Swahili is actually quite similar to the history of English, reflecting the movements and interactions of people over thousands of years. Swahili has a rich vocabulary with influences from various African tongues, Arabic, and some terms from European languages, Persian/Farsi, the Indian subcontinent, and beyond. The grammar is quite complex, so the language takes years to learn well.
The myth that Swahili is a simple or pidgin language is quite common in the US and Europe. Perhaps that is because of the Tarzan movies, where the Swahili used resembles the "me Tarzan, you Jane" quality of the English. Or perhaps it is because Swahili speakers tend to be very forgiving listeners, so visitors to East Africa get the feeling that they are communicating with just a few words of the language, because their hosts twist their ears in order to understand.
If you build it, they will come...
As to private donors: http://www.justgiving.com/pfp/swahili . So far, no dot.angel has emerged, though quite a few people have been extremely generous in helping keep the project going with relatively small donations.
Funding basically involves staffing, for programming and for editorial work. The more funding available, the more technically ambitious the project can be, and the more content we can provide. We would ideally like to expand the model to other languages, but, because the quality of the project demands scholarly oversight, we would need to actually hire people to work on additional languages and additional tasks. It's a case of getting what you pay for - the project aims to produce quality educational resources, which means that professional scholars need to give their time, and if they are giving the sort of time necessary to get the resources online this century, they need to be paid so they can buy food and pay the rent. Hosting costs are minimal - Yale is quite generous with server space. Publishing costs will be borne by the publishing house, when we eventually get to the point of producing print dictionaries, although we've got some problematic issues ahead because publishing houses are wary of printing something that is also available for free online. The project has a proven record of spending its money wisely and producing results, but it does need some sort of cash flow to keep doing the things it does!
If you build it, they will come...
Free text books means more money can be put into teacher salaries so we get the best and the brightest, and so children can have facilities that don't look like they've been abandoned for 25 years.
Abstinence is a government conspiracy. www.SafeSexZone.co