Universal Free Dictionary

Zdenek Broz writes "The all free dictionaries project focuses on maintaining free dictionaries (now more than 90 with more than 3,300,000 translations). We are designing a new system which will unite them all into one universal dictionary for all languages. The universal dictionary will be soon available for free under GPL."

Comments

Engrish Module?

[mfh]

This seems to be entirely useful. The proposed design looks like they will have a quick dictionary lookup on a word in the language being used with a definition, and cross-reference to the same word in other languages. That could be entirely useful, and anyone who enjoys Engrish might wish to help add that module (mostly for fun), but it looks like this project might actually take some of the mystery out of translation. Perhaps Engrish is going to be a thing of the past?

Re:Engrish Module?

[Anonymous Coward]

You might apperciate this.

I already have a pretty good dictionary

[teiresias]

I'm all for this but dictionary.com,Babelfish, and google meet my dictionary needs.

Re:I already have a pretty good dictionary

[jrl87]

i think the idea is to consolodate those down into one so you can use one site as apposed to three or how ever many you happen to use

Re:I already have a pretty good dictionary

[goon america]

There's also free/libre wordnet, wiktionary...

Why can't these projects work together? Seems like a lot of wheel-reinvention to me...

Why start a separate project?

[koreaman]

Why not just contribute to Wiktionary?
Or if they don't like the possibility of vandos, why not fork Wiktionary?

Re:Why start a separate project?

[dubl-u]

I think this is more like The Rosetta Project.

Urban Dictionary

[eln]

I think a multi-language Urban Dictionary for slang would be far more useful.

Re:Urban Dictionary

[gardyloo]

Perhaps. But the problem with slang in general, and "urban slang" especially, is that it is *so* dependent on puns, knowledge of popular (or geek) culture in one particular part of the world, and so forth. I fear that if many of these slang words/phrases are translated, unless the translators are especially good at capturing all or most of the "background" things in a given definition, the whole impact of the slang term will be totally lost. Explaining a joke usually takes the point of the joke and totally chews it up.

Somewhat like translating haikus into English. The whole 5-7-5 thing is fun and challenging, I suppose (I personally hated having to write them in middle school, mainly because it was in lieu of worthwhile reading and writing), but (supposedly; I don't know Japanese) the poems in the parent language probably have a lot of import that the translated-to language may lack.

Then again, a woman at a party once told James Thurber that she'd read a French translation of his My Life and Hard Times, adding, "You know, the book is even better in French!" To which Thurber replied, "Yes---my work tends to lose something in the original."

What?

[Anonymous Coward]

The GPL is a license for computer code. Don't you mean public domain, or the Creative Commons license?

why GPL?

[emkman]

Aren't there much better licenses for dictionaries than the GPL? Creative Commons comes to mind. What does the Guttenberg project use?

Re:why GPL?

[Anonymous Coward]

I agree with you. Though GPL can be used for any kind of work, it is best suited for works where is possible to identify a "source" and a "binary" (speaking in terms of software).

A good choice would be the GNU Free Documentation License if you ask me.

Using the GNU "Free" Documentation License is a bad idea.

phonetic transcriptions

[zobier]

Why don't more people put syllable boundary markers in their phonetic transcriptions?

Question about dictionaries under GPL license

[kurisuto]

Some of the texts on the Free Dictionaries project are listed as being licensed under the GPL. Can you mingle public-domain text with GPL'ed text?

This is a matter of practical concern. I'm overseeing a project which is digitizing copyright-expired dictionaries of the early Germanic languages. Some of the texts on my site are in German, and I'd like to use the GPL'ed Free Dictionaries German-English word list to add a feature to my project which allows you to click a German word to get a translation for that word.

Question 1: Are there provisions of the GPL which would prevent the a GPL'ed dictionary from being intermingled in this matter with existing public domain texts?

Another problem. The texts in my project contain many rare German words relating to Iron Age technology which are unlikely to be in the Free Dictionaries list, so I'd like to add my own supplemental list of words.

Question 2: Can I assign my supplemental word list to the public domain, or do I have to license it under the GPL as a modification to the original word list?

ideal for classrooms

[dazz_j]

I've been looking for something like this for the linux system in my wife's classroom. Not because of anything that special about the way definitions are developed, but just because it can be downloaded and used offline. (don't ask me why they can't run the network to the classroom, but the haven't). This could become one of the most popular programs there after Connect-4.

American Sign Language (AMSLAN?)

[buckhead_buddy]

Just today I was trying to prototype some ideas on a dictionary idea I had.

My idea was to create a tutorial for learning AMSLAN (American Sign Language) and to use texts from Project Gutenburg or other public domain works. One (of several) problems though is that English is rife with homographs... words that are spelled alike but have different meanings. In the case of sign language, a sign for a bow in a little girl's hair might appear extremely odd if the sign for bow of a ship popped up in automatic substitution.

Dealing with the homographs is a problem, but I see that this site's plan already takes a stab at dealing with such things (in their cat example). I'd love it if they went to the trouble of also including a bit of AMSLAN (either in animations or static pictures) as that might inspire me with some help in the solution.

Ideally, my desire is to get an automated library that could read a text (possibly read by the human sorting the homographs). And allow a user to listen to the reading and watch the text (while learning English), listen to the reading only (if hearing impaired), watch a silent sign language presentation with subtitles (to learn sign language), or watch a silent presentation through signing (if reading in silence is preferable).

Just kind of bizarre that the idea struck the same day as this article appeared, I thought. :-)

Dictionary != language textbook

[glasse]

Having taken Chinese I this term, I have learned that there's a whole lot more to a language than just vocabulary. In order to be a useful English-killer or monolinguism-killer, a language site needs to have information on how to pronounce words, how to write all of the glyphs used in the language (which might not be important if it uses a Latin-based script and so does your native language, but a lot of languages don't), and some idea of how to construct a useful sentence. (Word order, how to conjugate verbs/decline nouns, use of measure words/particles/prepositions, even synonyms and homonyms..) Also useful would be free media in the language -- TV shows, music, menus, newspapers -- but I know this would be very difficult to host effectively. My Chinese textbook's name translates as "Chinese Listening, Speaking, Reading and Writing". Developing skill with all of these facets of a language is part of gaining facility with it.

I would love a free-content languages database, full of audio samples of native speakers and grammar rules, but this isn't quite there yet. I do hope something like it gets off the ground, though, because monolinguism is a terrible disease in a global community :).

Ethan

Re:The flaw

[CRCulver]

It doesn't stomp out the very valuable native cultures of the world, it co-exists with them

No it doesn't. Try speaking any language besides Esperanto among Esperantists and see how quickly they complain. The hostility against learning and practising real languages in a fruitful and convenient international setting is what has driven me away from Esperanto. When I meet with, for example, Hungarian Esperantists, I would prefer to speak Hungarian with them, since I am already relatively proficient and further practise is always desirable. Esperanto should only be relied upon as a last result. But no, Esperantists insist on using only a single language, with no alternatives.

Furthermore, the idea that Esperanto is meant to "protect" minority language is a relatively recent sentiment, appearing in full force only with the Prague Manifesto of 1996. Before that Esperanto was, in the view of many, meant to supersede all national languages. Read the complete works of Zamenhof and see how often he espouses the hope that in the next several hundred years all people will have only a single language.

Maybe they could improve their English design?

[RedLaggedTeut]

Maybe they could improve their design by adding context to the english words, or maybe to ALL words in ANY language? For example (incorrectness possible, I'm not an arab speaker):

foundation/base(engineering,housebuilding,metaph or ical)
= al Qaeda(engineering,housebuilding,metaphorical)

the loo/the sit(colloq.,sanitary)=
al Qaeda(colloq.,sanitary)

a foundation(organisation,group)=
Al Qaeda(organisation,group)

Al Qaeda(terror,name)=
Al Qaeda(terror,name)

Re:There's more than this to a good dictionary

[Bill Walker]

If you'd like a good example of this, check out Whitaker's Words. It's a latin parser that does pretty much the morphological analysis you describe.

It's not perfect, principally because it doesn't have a convention to enter long vowels (the ones with a bar over them in latin), but it sure got me through High School latin.

That was back in '97 or so, and I see Mr.Whitaker is still updating the page, so maybe it's much better now. He was trying at one point to have the parsed latin translate directly into english; IIRC he was having more trouble constructing correct english than with decoding the latin.