I quote from your reference, Vilius - To translate accurately, a machine must be able to understand the text. It must be able to follow the author's argument, so it must have some ability to reason

This is patently not completely true.

Look how intelligently Google Translate handles the complications of the following:

Fruit flies like bananas
Time flies like an arrow
They like coffee
The Esperantists, like many before them, have asked the question

The French rendering is:

Les mouches des fruits comme les bananes
Le temps file comme une flèche
Ils aiment le café
Les espérantistes, comme d'autres avant eux, ont posé la question.

Not 100% perfect, but pretty impressive, in how it handles the ambiguity of 'flies' and 'like'

sudanglo:This is patently not completely true.

Unfortunately my French is not good enough to see what exactly happens in those sentences. Yet one correct translation of some trickier example is not enough to disprove my point that in some cases deep understanding of the context is required to make good translation. In the same article there is an example by Claude Piron:

[..] an Australian physician, cited the example of an epidemic which was declared during World War II in a "Japanese prisoner of war camp".

Now, who are the sick ones - Japanese, or some people captured and held as prisoners by the Japanese? Given enough context human translator would handle such ambiguity without many problems. Now try to create an algorithm capable of that, and you'll pretty soon find yourself on a very slippery slope towards strong-AI. And unfortunately our computers are not capable of doing that, and will not be any time in the near future.

marcuscf:

bartlett22183:

Tempodivalse:At least with EO, you know that once you learn a spelling/pronunciation rule, you'll never have to worry about exeptions.

Well, exactly the same principle is true, unambiguously and without doubt or hesitation, of Ido. In Ido, there are no exceptions. None. Period. It is just that the rules are different from those of Esperanto, and not any more burdensome in any meaningful way. ...

Well, nothing has exceptions if the rules are complex enough...

To me, the radio × omnadie problem (rádio × omna-díe with accents or radjo × omnadie with Esperanto spelling) brings enough doubt and hesitation during the learning process.

Different things for different individuals. It took me, oh, perhaps, thirty seconds to understand the Ido accentuation rules. On the other hand, I will probably to the end of my life mispronounce E-o 'radIo' simply because it does not correspond to my native (English) speech habits which I have never been able to overcome. Your experience may be different. I very much doubt that it is possible to construct any language useful for general communication which will have no difficulties for every adult learner. What is easy in Esperanto and difficult in Ido for one person may be just the opposite for another.

Vilius, Piron was probably writing before the advent of Google Translate, and before the Web had become firmly established, or before the number of Web users was substantial.

His view is therefore one of supposition of failure, not one based on fact.

So what does Google Translate make of your test phrase?

Here's the result in German:

in einem japanischen Kriegsgefangenenlager

Now I don't speak German so I don't know if Kriegsgefangenenlager is the right word for a prison of war camp. But it seems clear that the translation is of a prisoner of war camp run by the Japanese (not one containing Japanese prisoners), which as a native speaker of English I would say is the right interpretation.

The statistical approach that translation software like GT takes is only likely to give better results with the gathering of more bi-lingual texts, and the software also learns from human corrections.

In fact, any translation software does not need to have intelligence ascribed to it, but it may appear to be intelligent if is is teachable (ie can be told when it gets it wrong).

Vilius:And unfortunately our computers are not capable of doing that, and will not be any time in the near future.

Not that I disagree with you, but in my experience it's not always wise to say that "technology will never be able to accomplish such and such". 20 years ago it seemed impossible that any computer would ever surpass the human World Chess Champion. Today, a chess program running on an average-speed laptop can easily beat any chess player in the world.

Not to sound pessimistic, but I think Esperantujo should focus slightly less on the intercultural communication aspect of EO, and shift more focus on its propedeutic value. Many studies have already proven that learning EO prior to learning a "natural" second language results in the "natural" language coming much more easily later. If these results were promoted more, then perhaps Espranto would seem more attractive to those who fail to appreciate the more Zamenhof-esque idealism behind the language.

Although machine translation is nowhere near a level capable of fluently translating large texts, I've already gotten the reaction of "why Esperanto? I just use Google!" several times when chatting to other people about the benefits of learning EO. Public perception, I think, is going to lean increasingly in favour of machine translation and less towards an auxlang, and we need to anticipate that.

I am gald to be back at/with/in (who knows which prep is "right" here?) "lernu"! I was in the countryside in Brazil and the Internet connection was expensive and lousy. In the cities, Brasil is very modern - but - like other gigantic countries (USA included) the rural areas are less developed.
Any way, the Ido discussion is very lernu-like and gives me a chuckle. I once dabbled in learning a language that used German word roots with a simplified Germanic grammar - for Germanic language speakers it was great! It was a good way to limber up my "in-the-box" brain but not very useful for the rest of the world. The same goes for Ido (I speak several Romannce languages but they are not my favorite - though I think that I will end up living in Brazil - dang! love it! can't get Brazil off my mind).
Here is why Esperanto is sooooo cool: For the flight, I had on my "Mi parolas Esperanton!" shirt ($15 US from CafeExpress). In the line for the check-in, a woman in another line addressed me in Esperanto. She was def. not Brazilian and it sort of bonded us for a few short seconds. In front of the boarding gate a Brazilian woman who lives in Chicago asked me about Esperanto (she has lived 15 years in the US and is still struggling with Sudanglo's lingua franca). She had heard about E and was really excited about our language. Then in the airplane, waiting for my turn in the john, a Brazilian passenger told me that she was studying Esperanto but was too embarassed to try to speak. I told her, in Portuguese, that I have been struggling with Portuguese for 18 years and yet .... I still need a lot of struggling. Dang! I was so elated!
I suppose that someone may come up with simplified Turkic, Chino-Japanese or Bantu-without-borders - but E takes the cake!
Yeah, I write in somewhat unconventional E (since I speak Turkish a lot and like the Turkic kio-kioj words that complete the "tabelvortoj" for me. I am sure that E will not whither away because I commit this un-European "error" but - hey, wait a min! - people who speak E still understand me - and probably all Turks who might venture into Esperantio/Esperantujo/Esperantolando translating directly from Turkish. And that is awesome! Really, awesome, in both the old and new senses. Hand it to Z! He didn't shackle us - and that is how it should be in a lingua franca - no Given Pronunciation to pattern speech after, no Queen's (or Queens') English [sorry for the lousy pun but I live in Fort Lauderdale that is super gay]. Nobody like the people in the countryside of Brazil who didn't understand me because my Portuguese is not sterling - more like the urban Brazilians who "filled in the blanks" when I used a non-standard form. So do that Ido! Give me a break and let me free! Esperanto estas mia amo! The dizzying country air but urban street-smart.

Oh, and one more note: I know that I make mistakes when writing English fast like this - and I wrote my friggin' dissertation in the language - but I will NEVER be perfect with English spelling - just don't have the time to devote to it since there are so many other beautiful things in the world to learn. Chaulk it up for E! Spelling is easier!

robbkvasnak:Really, awesome, in both the old and new senses. Hand it to Z! He didn't shackle us - and that is how it should be in a lingua franca - no Given Pronunciation to pattern speech after, no Queen's (or Queens') English [sorry for the lousy pun but I live in Fort Lauderdale that is super gay]. Nobody like the people in the countryside of Brazil who didn't understand me because my Portuguese is not sterling - more like the urban Brazilians who "filled in the blanks" when I used a non-standard form. So do that Ido! Give me a break and let me free! Esperanto estas mia amo! The dizzying country air but urban street-smart.

What does this have to do with the topic of this thread ("Why not Ido?" )? Is there some particular reason why you think Esperanto is better than Ido?

sudanglo:The statistical approach that translation software like GT takes is only likely to give better results with the gathering of more bi-lingual texts

You may have infinite amount of texts relating to the World War II, yet no statistical approach will be able to reliably figure out who are the prisoners. Only deep understanding of context can help to sort that out. And deep understanding of context, by definition, is strong-AI.

sudanglo:In fact, any translation software does not need to have intelligence ascribed to it, but it may appear to be intelligent if is is teachable (ie can be told when it gets it wrong).

There is nothing inherently intelligent in us humans as well, we also just appear to be intelligent. My point is that computers will not reach our level of apparent intellect in foreseeable future. Even if they would, we'd likely have much better uses for such machines, than translation (I'd prefer to speak to such a computer itself, not use it as a mere translator between myself and another boring human). Therefore I don't believe that automatic translation is, or will be a threat to Esperanto, as you have stated earlier in this thread.

Tempodivalse:Today, a chess program running on an average-speed laptop can easily beat any chess player in the world.

(Un)fortunately, understanding human utterance is by orders of magnitude more difficult than playing chess (it's just a game after all). Next time you read some non-trivial text pay attention to all those twisted analogies, ironies, and loads of other literary devices, we use to carry through all those layers of communication. Also note all the details about our world we do not write/say explicitly, but just expect our reader to know in advance. Being able to understand all that and render it into another human language without losing too much of the nuance is a task for a system just as intelligent as us.