Language desk
< September 24	<< Aug \| September \| Oct >>	September 26 >

Welcome to the Wikipedia Language Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

September 25

Machine translation

Are free products like the online Google and Microsoft translators state of the art, or are paid-for programs a lot better? 86.186.35.26 (talk) 02:02, 25 September 2010 (UTC).[reply]

It would help if you could give an example of what 'paid for' programs you are referring to, as there are many many translation programs out there, many of which serve a specific purpose. --KägeTorä - (影虎) (TALK) 02:22, 25 September 2010 (UTC)[reply]

I'm interested in where the state of the art currently is, not in an evaluation of particular commercially available products, so I don't have any specific examples in mind. Really what I am asking is "for general purpose translation, is Google/Microsoft/Babel Fish as good as it gets given current technology?" —Preceding unsigned comment added by 86.186.35.26 (talk) 02:32, 25 September 2010 (UTC)[reply]

Google Translate and Yahoo! Babelfish are pretty much the best automatic services out there, yes. However, if you want a truly "state-of-the-art" translation, your best choice is always to ask a real person who speaks the language, whether through an online paid service or otherwise. There are a lot of people within the Wiki-projects who speak various languages that might do it for you free, also. Lexicografía (talk) 02:43, 25 September 2010 (UTC)[reply]

Thanks, yes, I am aware that good human translators far exceed the capabilities of machine translation. What I am not so clear about is whether the best machine translators exceed the free products I mentioned. From what you say, it sounds like they don't. 86.186.35.26 (talk) 03:04, 25 September 2010 (UTC)[reply]

All machine translators have problems and shouldn't be completely trusted (for instance, not too long ago Google Translate translated "assemble a trivia quiz" into Spanish as "form a bond with Israel"). What languages are you translating from and into? Lexicografía (talk) 03:09, 25 September 2010 (UTC)[reply]

I do not have any specific translation task in mind. My question is about the present capabilities of translation technology. I have casual experience of translation into English from a variety of languages using all the free products I have mentioned, and I am well aware that they all readily produce garbage. 86.186.35.26 (talk) 03:23, 25 September 2010 (UTC).[reply]

"Paid" translation programs usually serve a specific purpose - setting them apart from the free stuff. It's better at its one specific purpose, but worse at general purpose translation. So I'm told, anyway. 90.193.232.118 (talk) 11:31, 25 September 2010 (UTC)[reply]

OP, you haven't clarified what you mean by 'better'. 'Better' for what? Anyway, the simple answer is 'no'. They are not a lot better. Reason being that none of them, paid for or not, is as reliable or as good as a human translator. Sure, some of these softwares might have little features like 'translate into colloquial language' or 'use medical terms' or 'keep to mechanical engineering terminology' or whatever, but they still get the stuff wrong a lot of the time. If you're just using it for casual stuff, then it's totally OK to use machine translation - it can be great fun unravelling the mess afterwards - but for industrial standard translation, it is not advisable (purposefully understated). --KägeTorä - (影虎) (TALK) 12:28, 25 September 2010 (UTC)[reply]

I've said that I am asking about "general-purpose translation". Other than that, I'm afraid I cannot explain what I mean by "better" except by stating the obvious. 86.174.167.79 (talk) 14:23, 25 September 2010 (UTC).[reply]

Right, I'm not trying to be funny, but I'm not sure how I can put it more clearly. I and others have been trying to tell you that the answer is 'no, they are not much/any better'. They aren't. On a side note, you may need to 'state the obvious', because what may be 'obvious' to you may not be obvious to others. It is clear that you are not getting the answer that you want, and this must mean that there is a difference in how we understand your question and how you understand it. --KägeTorä - (影虎) (TALK) 15:00, 25 September 2010 (UTC)[reply]

Are there any paid-for translation programs which aren't made for a specific purpose? If not, the question is irrelevant as the circumstances to which the OP refers don't exist. Vimescarrot (talk) 16:33, 25 September 2010 (UTC)[reply]

Well, before the free stuff came out there used to be a lot of them. I suspect that since we can now get gibberish for free on Google, companies have stopped trying to make us pay for it, unless somehow they can market their software as something special and not just generic. --KägeTorä - (影虎) (TALK) 16:42, 25 September 2010 (UTC)[reply]

@KageTora. You seem to be frustrated that you "are not sure how you can put it more clearly", and you talk about what you have been "trying to tell me", but prior to the post in question you had not put it in any way at all or told me anything, you had simply asked for clarification which I tried to provide. In the post in question you offered an answer but nevertheless again asked for clarification, so I tried to respond to your request, which you can hardly object to. 86.186.35.223 (talk) 17:41, 25 September 2010 (UTC).[reply]

Sorry, OP, but since when did "the simple answer is 'no'. They are not a lot better" not mean exactly what it says? This is what I was referring to, in case you missed it. I am really sorry, but this seems to be devolving into a kind of argument, which was not my (nor your) intention. --KägeTorä - (影虎) (TALK) 18:30, 25 September 2010 (UTC)[reply]

I understood your post perfectly well. As part of your post, you asked me what I meant by "better", and I responded to that question. Then, for some reason that completely eludes me, you start talking to me as if you've had to explain things over and over and I'm some kind of retard for not understanding you. I don't want to argue with you either, and I do not understand how we have reached this position. I propose that we now let it rest. 86.186.35.223 (talk) 19:02, 25 September 2010 (UTC)[reply]

Understood. I hope we can talk about something else some other time. --KägeTorä - (影虎) (TALK) 19:10, 25 September 2010 (UTC)[reply]

@86.186. Has your question been answered, or is there anything you are still seeking to know? If there is, what is it? Lexicografía (talk) 17:47, 25 September 2010 (UTC)[reply]

I don't know why you 2 (Kage and Lexi) couldn't understand the question, it is quite plain to me. Unfortunately I am not the owner of any licence for paid translation program or service, so I cannot answer. The question is: are the free AUTOMATED translation service significantly less accurate for genmeral purpose translation than any paid AUTOMATED translation service or software. If you have never used a paid AUTOMATED translation service, don't bother answering, you will not know the answer. If you have used a paid AUTOMATED translation service, let us know which one and what you thought of its accuracy, so we may have some idea of the value of your opinion. --Lgriot (talk) 13:11, 27 September 2010 (UTC)[reply]

I used to use one about 5 or 6 years ago, called The翻訳 v.7.0 (by Toshiba, I think), and it was not considerably better than any of the free services we have. --KägeTorä - (影虎) (TALK) 13:42, 27 September 2010 (UTC)[reply]

Variety of English taught in China

According to my parents, the variety of English that has been traditionally taught in China is British English, because of China's economic ties with the UK during the 19th and early 20th centuries. Is this accurate? Is British English still the standard taught in China, or have things shifted toward American English? 98.116.90.160 (talk) 07:06, 25 September 2010 (UTC)[reply]

In my experience it tends to vary from school to school - just depends on where the teacher is from. There are national exams, and they seem to be focused on British English, though, if that's what you want to know. --KägeTorä - (影虎) (TALK) 12:11, 25 September 2010 (UTC)[reply]

Like Kage says, it varies, but in my experience it's shifting to AmEng these days. Most Chinese people my age (university students and grad students) speak with British accents (if they know enough English to speak it at all), but little kids more often have American accents (again, if they know enough to have any discernible accent at all). rʨanaɢ (talk) 12:14, 25 September 2010 (UTC)[reply]

My impression is that English is so difficult for native Chinese speakers that the Chinese accent almost always overwhelms the more subtle differences between British and American English. Looie496 (talk) 17:40, 25 September 2010 (UTC)[reply]

Not necessarily. Actually - and this is still relevant to the OP's question and I am surprised I didn't remember it before - when I lived in China way back in 1993, there was a girl I met and worked with on a magazine. She was a native Chinese and had never been outside of China, but spoke with the strongest Georgian accent (as in the US state, not the country near the Black Sea) you have ever heard, so much so that she was proud of it. She'd spent years trying to perfect this accent. In fact, she was so proud of this accent and held it in such value that very often she told me to shut up, lest my British accent contaminate hers. When I was in Shanghai last year, on the rare occasions that I met anyone who spoke English to me, they generally had a mixture of British and American accents. --KägeTorä - (影虎) (TALK) 20:31, 25 September 2010 (UTC)[reply]

There are a few quick & dirty diagnostics I sometimes use when I meet a Chinese person to figure out which English variety they were educated in. One is asking them to spell words like "color", "flavor", and "center", although if their English is good then these might be too obvious (they might be pretty aware of the differences already). Another is asking them to pronounce "Harry Potter"—works every time. rʨanaɢ (talk) 21:53, 25 September 2010 (UTC)[reply]

@Lookie: Since most North American accents are rhotic and most British accents are not (including the accents best known outside their respective territories, in both cases), English spoken by a foreigner usually quickly strikes me, at least, as British or Americanh, even where it is overlaid by a strong foreign accent - even Chinese. --ColinFine (talk) 18:51, 26 September 2010 (UTC)[reply]

Mildly relevant, I've heard (and it doesn't really surprise me) that often an English teacher in private institutions in CHina generally has to be white to meet parents approval. So a non-native white speaker, e.g. a German may be hired even if their English isn't that good. But a native English speaker of some other ethnicity (particularly Chinese), even with excellent English and from New Zealand, US, UK, Ireland, Canada, Australia etc; will not (or at least not in preference to the worse white speaker). So you may get a few Chinese with German, French, Dutch, or whatever accents, but probably not many with Malaysian, Indian, Singaporean, etc accents Nil Einne (talk) 21:32, 26 September 2010 (UTC)[reply]

That may be the case in relatively privileged settings (private tutors for wealthy people, top institutions, etc.), but a lot of places, like language training centers (培训中心) offering evening classes and stuff, will take whatever they can get. A few training centers I worked at in Beijing and Ürümchi had no non-Chinese teachers other than me; all the other teachers were Chinese people with greater or lesser English proficiency... rʨanaɢ (talk) 21:42, 26 September 2010 (UTC)[reply]

Perhaps I explained it poorly. I think the point of the comments I've read is that for a number of private institutions there is demand for white teachers. Obviously quite a few places do have Chinese, primarily native Chinese but their pay is generally very poor by international standards, so most people from other countries, particularly developed ones aren't particularly interested in working for them. (There would also be I presume visa issues with hiring a non Chinese citizen.) And in any case, when it comes down to it, a 'white' teacher even one with clearly less expensive and a lower level of English will often get preference to a non-white. See [1] [2] for example (although the second one suggests even being a highly competent and experienced non-native 'white' speaker can be just as problematic). The second one links to a few other discussions and there are plenty of comments in other places. It's of course difficult to know how widespread these practices actually are and how much they are just people who like to think of China (and other similar places) as 'odd' or 'bad'. Nil Einne (talk) 13:58, 27 September 2010 (UTC)[reply]

I searched this out a long time later to check something and just wanted to add a key issue is probably how their pay varies depending on their perceived 'race'. It may be an Chinese American can expect significantly less even if they e.g. actually have TESL training and obviously an excellent native level of English and gets paid perhaps closer to a native Chinese person with far poorer English and credential, compared to a 'white' American with the same level of English as the Chinese Americans. Nil Einne (talk) 09:59, 23 December 2012 (UTC)[reply]

I would say it varies with regions. In Beijing (where I live and study) it's British English, though American English is gaining popularity amongst students. Most Chinese students have a strong tendency towards American pronunciation and will do anything to achieve it, even though the sounds are so overwhelmed with typical Chinese intonations that they sound dreadful.Faiz aguo 01:30, 29 September 2010 (UTC)[reply]

English language in Quebec

QUOTE: Percentage of population(Quebec) that is fluent in French (95.0%)
Question: What is the percentage of Quebec population that is fluent in English? I I would expect it is more then 50%, but could not find any definitive statistics... 76.67.10.16 (talk) 09:32, 25 September 2010 (UTC)[reply]

That article seems to be mostly unsourced, but if you take its numbers you can do the math:

12.6% (according to the article) are native anglophones
34% of native francophones (who are 65% of the overall population) also speak English; 34% of 65% is 22%
19% of the "allophones" (native speakers of langs. other than English or French), who themselves are20.4% of the population, also speak English. 19% of 20.4% should be about 4%.
Adding those up, you get about 38%. Sounds low to me, but like I said that article is mostly unsourced. rʨanaɢ (talk) 12:11, 25 September 2010 (UTC)[reply]

Based on some time spent in Quebec, 38% sounds about right to me—if not a bit high—for fluency in English. Fluency is more than knowing a few words and phrases or being able to communicate in a halting way, which I think a majority of Quebecois can do in English. Fluency entails a high degree of proficiency in English, which I think is beyond the abilities of the majority in Quebec. If you confine yourself to the main centers for business and tourism in Montreal and the neighborhoods where Anglophones live, it is true that most people you encounter will be fluent in English. However, those areas make up a small minority of the population of the province. Outside of central and western Montreal, only a minority are truly fluent in English. Marco polo (talk) 13:25, 25 September 2010 (UTC)[reply]

That's a good point. Most of the time I have spent in Quebec was in the major cities, and even there most of my interactions in English were quite simple, not enough to judge whether people were really fluent or just had some everyday phrases. rʨanaɢ (talk) 13:39, 25 September 2010 (UTC)[reply]

French Food

I notice something in a French butchers that looked like liver or pâté. It had a name along the lines of foit du guissé, or something like that. Does anyone have any idea of what it was really called, and what exactly it was? Thanks in advance. — Fly by Night (talk) 11:42, 25 September 2010 (UTC)[reply]

Foie gras? ---Sluzzelin talk 11:43, 25 September 2010 (UTC)[reply]

No, it's certainly not foie gras. I know that about that stuff already. I was thinking that this stuff might be the same, but from a different animal. So maybe it said foie guissé? — Fly by Night (talk) 12:26, 25 September 2010 (UTC)[reply]

There's pâté de foie gras...Lexicografía (talk) 12:37, 25 September 2010 (UTC)[reply]

The closest I can think of is foie de génisse, which is cow's liver. Deor (talk) 13:15, 25 September 2010 (UTC)[reply]

There are a lot of similar French Pâtés, e.g. Pâté de campagne (mostly pork), Pâté de canard, and so on, many aromatised with fruits or fruit brandies. --Stephan Schulz (talk) 15:44, 25 September 2010 (UTC)[reply]

May be foie déguisé (disguised liver), if it's a speciality of this butcher (I have never heard of such pâté). This phrase could have been coined from fruits déguisé [3]. — AldoSyrt (talk) 17:44, 25 September 2010 (UTC)[reply]

But disguised as what? There could be an unpleasant surprise for some poor unsuspecting customer.... rʨanaɢ (talk) 21:58, 25 September 2010 (UTC)[reply]

I think that Deor has got it. Foie de génisse is only one letter off what I thought I had seen; give or take a misplaced accent. Thanks Deor! — Fly by Night (talk) 18:40, 25 September 2010 (UTC)[reply]

Lithuanian?

If anyone speaks Lithuanian, they might be able to help with this thread on the Misc. desk. I'm not sure how clearly the Original Poster and the answerers are communicating with each other! 109.155.33.219 (talk) 12:54, 25 September 2010 (UTC)[reply]

It looks like they are communicating perfectly well in English. --KägeTorä - (影虎) (TALK) 14:14, 25 September 2010 (UTC)[reply]

Oh good. I was concerned that subtleties might not be understood, and that the Google-translated sentences might muddy the waters. But if other people think it's fine, it probably is. 109.155.33.219 (talk) 15:16, 25 September 2010 (UTC)[reply]

"Luke"

What is the origin of "luke" to mean "somewhat" in lukewarm (at least "somewhat" or "mildly" is what I think it adds to "warm"). Was lukecold ever an expression? And are there any other words that also use the "luke" prefix in combination? Just curious. Thanks.--162.84.161.15 (talk) 17:14, 25 September 2010 (UTC)[reply]

First off, the "luke" in "lukewarm" doesn't mean "somewhat", it means "warm" as well, so it's a kind of redundant word. But as for the etymology, it's originally English — it's from a Middle English word, spelled either luke or lewk (which survives today only in some British dialects), from Old English hleow meaning warm. The modern word lee, as in the sheltered side of something, comes from the same Old English root. Lexicografía (talk) 17:37, 25 September 2010 (UTC)[reply]

The German adjectives lau and lauwarm both mean "lukewarm".—Wavelength (talk) 18:21, 25 September 2010 (UTC)[reply]

See http://www.etymonline.com/index.php?search=lukewarm&searchmode=none. -- Wavelength (talk) 18:24, 25 September 2010 (UTC)[reply]

I agree with the OP that lukewarm water is somewhere between cold and warm without qualifications, i.e somewhat warm. Whatever the origins of luke- may be, not even etymologists mean "warm warm" when they say "lukewarm". -- Jack of Oz ... speak! ... 19:32, 25 September 2010 (UTC)[reply]

Indeed. The word does not mean "very warm", as its etymology might suggest, it just means "mildly warm". Lexicografía (talk) 19:36, 25 September 2010 (UTC)[reply]

It is worth checking out Wavelength's link. The origins don't seem to be clear. Either Middle English leuk for "tepid", perhaps from Middle Dutch or Old Frisian "leuk". Or from the Old English word hleowe for "warm", as mentioned by Lexicografia. ---Sluzzelin talk 20:08, 25 September 2010 (UTC)[reply]

Similarly, A Concise Etymological Dictionary of the English Language traces it back to the Early Frisian word for "tepid, weak, slack".[4] ---Sluzzelin talk 20:14, 25 September 2010 (UTC)[reply]

I only gave the etymology as explained in Merriam-Webster's dictionary, which only gives the story of it being through Old English. Lexicografía (talk) 20:24, 25 September 2010 (UTC)[reply]

As has been pointed out above, a similar term, lauwarm (=lukewarm) exists in German. It may be of interest to note that the adjective lau ("luke") and the noun Lauheit ("lukeness") do exist in the German language. Both are used infrequently and often describe a sort of wishy-washy / undecided stance on some matter. The only common (albeit old-fashioned or jokingly poetic) usage of lau is in describing the air-temperature of a given day, when it is the equivalent to "mild" / "balmy". --Cookatoo.ergo.ZooM (talk) 21:18, 25 September 2010 (UTC)[reply]

Off-topic ~~argument~~ linguistic debate between Lexicografía and JackofOz

Sorry to quibble, but which of your posts do you want us to accept, Lex? Your first one denied that it means "somewhat". But this one says it means "mildly". In this context, "somewhat" and "mildly" mean exactly the same thing, do they not? -- Jack of Oz ... speak! ... 19:58, 25 September 2010 (UTC)[reply]

Sorry if I was confusing. I only meant that the "luke-" element does not mean "somewhat", as the OP was suggesting. Lexicografía (talk) 20:00, 25 September 2010 (UTC)[reply]

There's a difference between etymology and current meaning. You said above "I only meant that the "luke-" element does not mean "somewhat", as the OP was suggesting". Not in the sense that you can apply it to any old adjective (e.g. "green") to make it mean "somewhat green". No, that's true, you can't do that. It's only ever used in the word "lukewarm", as far as I know. But in that word, whatever it may originally have meant, it does now mean "somewhat", "moderately" or "mildly". That's what people mean when they say "lukewarm". Is this not so? -- Jack of Oz ... speak! ... 20:31, 25 September 2010 (UTC)[reply]

The morpheme luke- does not, nor has ever meant, somewhat. The word lukewarm has, and has always meant, somewhat warm. Are you saying that those two facts cannot coexist? Lexicografía (talk) 20:33, 25 September 2010 (UTC)[reply]

Hypothetically, yes. If luke- had always had a wider application than it does, it would probably generally mean "warm(ly)" or "tepid(ly)". But it has never had that wide application, so its hypothetical meaning is neither here nor there, and there's no point reserving that meaning for the day when people might start using it to mean that. Language doesn't work that way. Ultimately, words mean what the users decide they mean, and if that's different from what the lexicians and the lexologists say, so be it, they have to duly record these meanings in their learned tomes. "Lukewarm" is the only place in the entire language where luke- is ever used. In that word, it means "somewhat/mildly". Effectively, then, luke- has changed its meaning from "warm" to "somewhat/mildly". If anyone ever did start to use new luke- words (such as lukehappy or lukeinterested or lukeargumentative), they would expect others to understand them as a qualification of the head word in exactly the same way in which luke- applies in "lukewarm". That is, "somewhat", "mildly" or "moderately". So we have a tension between a hypothetical that might have been but never was, and a hypothetical that could yet happen (which, given the propensity for people to come up with new jargon these days, is more likely than not). -- Jack of Oz ... speak! ... 21:10, 25 September 2010 (UTC)[reply]

OK, I surrender, you're right. luke- seems to have taken on a new meaning as "somewhat". I was going to use the example of tomboy, where the element Tom, meaning boy, wasn't used in other words, but it seems I'm wrong on that one too. However, even though the (misanalyzed) meaning of "luke" in words such as "lukecold" is somewhat, I still am convinced the meaning of "luke" in "lukewarm" is not somewhat, but warm/tepid. Yes? Lexicografía (talk) 21:20, 25 September 2010 (UTC)[reply]

Well, as I've tried to explain above, No. That's what it originally meant, centuries ago. It doesn't mean that anymore. It just doesn't. I can see we're going to have to agree to disagree on this. -- -- Jack of Oz ... speak! ... 21:31, 25 September 2010 (UTC)[reply]

Ok, I can do that. Lexicografía (talk) 21:41, 25 September 2010 (UTC)[reply]

The form luke in the word lukewarm is a cranberry morpheme. -- Wavelength (talk) 02:09, 26 September 2010 (UTC)[reply]

Sweet. I knew there had to be a word for this... Lexicografía (talk) 12:12, 26 September 2010 (UTC)[reply]

The OED also suggest the Old English "hléow" as the origin, and this view might be supported by the old dialectal "lew-warm" (tepid). I suspect that the Lancashire usage of "lew" (weak, poor, diluted) is derived from this, though the OED doesn't seem to have noticed. Dbfirs 03:08, 26 September 2010 (UTC)[reply]

The element quasi- means "somewhat". See http://www.etymonline.com/index.php?search=quasi-&searchmode=none.

—Wavelength (talk) 15:32, 26 September 2010 (UTC)[reply]

In Norwegian we simply say lunken; probably the same word as in English and German but without the redundancy? 62.16.146.27 (talk) 17:59, 26 September 2010 (UTC)[reply]

I can't resist repeating the old joke about asking a Yorkshireman how cold the water is in the swimming pool and being told "lukewarm". After jumping-in to find that the water is freezing cold, the Yorkshireman is questioned about his advice; "Well, it luked warm to me" comes the reply. Boom, boom! Alansplodge (talk) 21:59, 26 September 2010 (UTC)[reply]

The Oxford English Dictionary says "luke" means "lukewarm," as in this 14th century citation from Richard Rolle: "Als a lewke bath nouther hate ne calde." The etymology is unclear but may come from the dialect word "lew," which means "weak" or "pale." That word, in turn, is "of obscure origin." -- Mwalcoff (talk) 02:34, 27 September 2010 (UTC)[reply]

On the contrary, I think that the Lancashire usage of "lew" meaning weak or pale is a development from the much earlier usage of "lew" from "hléow". I would find it difficult to prove this, though. Dbfirs 18:23, 27 September 2010 (UTC)[reply]