Wikipedia:Reference desk/Archives/Language/2022 September 8
Language desk | ||
---|---|---|
< September 7 | << Aug | September | Oct >> | September 9 > |
Welcome to the Wikipedia Language Reference Desk Archives |
---|
The page you are currently viewing is a transcluded 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 8
[edit]"ungrammatical nonsense"?
[edit]What is wrong with the following sentence?
- Modern diagnostic equipment no longer needs this highly risky procedure.
I do not see any grammatical error here, while some user called it "ungrammatical nonsense". Any ideas? 85.193.215.210 (talk) 00:27, 8 September 2022 (UTC)
- It's a trifle informal, and I'm uncomfortable attributing "needs" to equipment. On the whole, the edit is good but the edit summary was unnecessarily brusque and snarky. --Orange Mike | Talk 00:48, 8 September 2022 (UTC)
- The sentence as quoted could make sense, but the IP omitted the context. There was a procedure to test for the presence of beryllium. Modern equipment is now used instead of that procedure. Saying that modern equipment "no longer needs" the procedure it superseded is indeed ungrammatical nonsense. Kzqj (talk) 11:58, 8 September 2022 (UTC)
- It's not ungrammatical, but it is illogical and nonsensical: the difference is important. {The poster formerly known as 87.81.230.195} 90.193.131.160 (talk) 12:44, 8 September 2022 (UTC)
- "Necessitates" could perhaps be used instead of "needs" to convey the meaning described by Kzqj. Hassocks5489 (Floreat Hova!) 12:45, 8 September 2022 (UTC) – Or, on reflection, it could perhaps be restated as Modern diagonstic equipment has rendered this highly risky procedure unnecessary (or "...renders this...") Hassocks5489 (Floreat Hova!) 13:25, 8 September 2022 (UTC)
- It was "necessitates" until two weeks ago, when an IP from Poland replaced "necessitates" by "needs" and again one week ago per ongoing edit war. --Pp.paul.4 (talk) 13:54, 8 September 2022 (UTC)
- "needs" is vastly better than "necessitates". I find it hard to imagine any sentence in which "necissitates" would be an improvement on "needs". DuncanHill (talk) 18:15, 8 September 2022 (UTC)
- I admit I had not read the sentence in its proper context (with the preceding sentence about scientists tasting beryllium) until now. Having now read it, and thinking about how I would write it: I think I would have gone with my second sentence ("Modern diagonstic equipment has rendered this highly risky procedure unnecessary"). Hassocks5489 (Floreat Hova!) 20:58, 8 September 2022 (UTC)
- "needs" is vastly better than "necessitates". I find it hard to imagine any sentence in which "necissitates" would be an improvement on "needs". DuncanHill (talk) 18:15, 8 September 2022 (UTC)
- It was "necessitates" until two weeks ago, when an IP from Poland replaced "necessitates" by "needs" and again one week ago per ongoing edit war. --Pp.paul.4 (talk) 13:54, 8 September 2022 (UTC)
- "Necessitates" could perhaps be used instead of "needs" to convey the meaning described by Kzqj. Hassocks5489 (Floreat Hova!) 12:45, 8 September 2022 (UTC) – Or, on reflection, it could perhaps be restated as Modern diagonstic equipment has rendered this highly risky procedure unnecessary (or "...renders this...") Hassocks5489 (Floreat Hova!) 13:25, 8 September 2022 (UTC)
- It's not ungrammatical, but it is illogical and nonsensical: the difference is important. {The poster formerly known as 87.81.230.195} 90.193.131.160 (talk) 12:44, 8 September 2022 (UTC)
- The sentence as quoted could make sense, but the IP omitted the context. There was a procedure to test for the presence of beryllium. Modern equipment is now used instead of that procedure. Saying that modern equipment "no longer needs" the procedure it superseded is indeed ungrammatical nonsense. Kzqj (talk) 11:58, 8 September 2022 (UTC)
- Per the above, while slightly informal (such as anthropomorphizing "diagnostic equipment" by using the verb "to need" which implies intent), there's is not much wrong with the sentence. If one wanted to avoid the anthropomorphism problem with "need", one could rewrite it in passive voice "It is not necessary to use this highly risky procedure with Modern diagnostic equipment" is quite a bit longer, but avoids the anthropomorphism problem. I don't find anything at all illogical, nonsensical, or ungrammatical with the OP's sentence, except in a strictly pedantic way; but given the fact that natural language is often different from formal standard language, it's not hard to understand in any way. --Jayron32 12:50, 8 September 2022 (UTC)
- Quibble: The sentence It is not necessary to use this highly risky procedure with Modern diagnostic equipment does not in fact use passive voice. A rewording in passive voice might be This highly risky procedure need not be used with modern diagnostic equipment. I'm not quite sure what the sentence means, in either form (modern diagnostic equipment removes the need, or the procedure does not require the modern equipment?) but the syntactical distinction can be made without figuring that out. --Trovatore (talk) 19:37, 8 September 2022 (UTC)
- The statement implies that modern equipment previously "needed" the risky procedure. That is nonsense. The procedure is no longer needed, because modern equipment exists. "Need" has no agent here; giving it one makes the sentence ungrammatical nonsense. Kzqj (talk) 13:12, 8 September 2022 (UTC)
- Grammar is about the way words function, every word in the sentence is using its proper function. "Colorless green ideas sleep furiously" is perfect grammar, even if it makes no sense otherwise. If the sentence is factually wrong, then it is factually wrong. It is not, as written, in any way ungrammatical. It uses nouns, verbs, and adjectives and other forms of English syntax and sentence structure entirely correctly. --Jayron32 15:15, 8 September 2022 (UTC)
- Most linguists do not regard a well-formed sentence that breaks semantic constraints as "ungrammatical". ColinFine (talk) 15:16, 8 September 2022 (UTC)
- I honestly don't think that arguing about whether this sentence is grammatically incorrect nonsense or just nonsense is worth anyone's time. I note that the Polish IP that started this thread has posted in several locations recently with silly questions about straightforward English usage ([1],[2],[3]). It looks like trolling to me. Kzqj (talk) 17:40, 8 September 2022 (UTC)
- @Kzqj This is irrelevant for the current topic. However, just because you don't understand something, doesn't make it silly. For native speakers almost everything in their language seems natural and logical, even
at the expense of[if it leads to] ambiguity and miscommunication. This is because they acquire their language mindlessly, which is acceptable in the early childhood, but editors of an encyclopedia should be more aware. My approach, even to my native language, is different because I don't want to be like a fish who does not see the water in which it swims. 85.193.215.210 (talk) 15:55, 9 September 2022 (UTC)
- @Kzqj This is irrelevant for the current topic. However, just because you don't understand something, doesn't make it silly. For native speakers almost everything in their language seems natural and logical, even
- I honestly don't think that arguing about whether this sentence is grammatically incorrect nonsense or just nonsense is worth anyone's time. I note that the Polish IP that started this thread has posted in several locations recently with silly questions about straightforward English usage ([1],[2],[3]). It looks like trolling to me. Kzqj (talk) 17:40, 8 September 2022 (UTC)
- It's neither ungrammatical nor nonsense; the meaning is quite clear. It's slightly metaphorical but (as DuncanHill mentions above) clearly superior to "necessitates". It's arguably not in an encyclopedic register; personally it's not far enough off to bother me, but since it bothers some people, swap it out for "requires" and be done with it — hard to see how anyone can argue with that. --Trovatore (talk) 19:48, 8 September 2022 (UTC)
- It is nonsense, and if "needs" were switched for "requires", it still would be. It is like saying "cars no longer need horses", when what you are trying to say is that because cars were invented, people don't use horses for transport any more. "Cars no longer require horses" would obviously not be any better. You can only read the sentence as non-nonsensical if you interpret the risky procedure as something that was done to old equipment but is no longer done to modern equipment - analogous to "cars no longer need crank-starting" - but that is not the intended meaning. Kzqj (talk) 19:59, 8 September 2022 (UTC)
- @Kzqj Suddenly, you write "It is nonsense", not "ungrammatical nonsense". Did you start to understand the word "grammar"? I ask you seriously because if this is the case should I understand that my wording is no longer ungrammatical nonsense, but just nonsense, though the grammar is correct? Of course, equipment does not have needs or desires, at least not at present, but we can say that "non-electric cars need fuel" in the sense that "fuel is necessary for them to function". Maslow's hierarchy of needs does not apply here. Note that in Google Search the phrase "cars need fuel" returns 22k hits (normally) and 1.5k hits in Google Books. But I admit that the use of the word "need" is a bit risky here. 85.193.215.210 (talk) 16:35, 9 September 2022 (UTC)
- You have not understood the meaning of the words in the sentence. You are evidently not capable of understanding them. Do you know the word "vexatious"? Kzqj (talk) 18:47, 9 September 2022 (UTC)
- @Kzqj Suddenly, you write "It is nonsense", not "ungrammatical nonsense". Did you start to understand the word "grammar"? I ask you seriously because if this is the case should I understand that my wording is no longer ungrammatical nonsense, but just nonsense, though the grammar is correct? Of course, equipment does not have needs or desires, at least not at present, but we can say that "non-electric cars need fuel" in the sense that "fuel is necessary for them to function". Maslow's hierarchy of needs does not apply here. Note that in Google Search the phrase "cars need fuel" returns 22k hits (normally) and 1.5k hits in Google Books. But I admit that the use of the word "need" is a bit risky here. 85.193.215.210 (talk) 16:35, 9 September 2022 (UTC)
- It is nonsense, and if "needs" were switched for "requires", it still would be. It is like saying "cars no longer need horses", when what you are trying to say is that because cars were invented, people don't use horses for transport any more. "Cars no longer require horses" would obviously not be any better. You can only read the sentence as non-nonsensical if you interpret the risky procedure as something that was done to old equipment but is no longer done to modern equipment - analogous to "cars no longer need crank-starting" - but that is not the intended meaning. Kzqj (talk) 19:59, 8 September 2022 (UTC)
- It's neither ungrammatical nor nonsense; the meaning is quite clear. It's slightly metaphorical but (as DuncanHill mentions above) clearly superior to "necessitates". It's arguably not in an encyclopedic register; personally it's not far enough off to bother me, but since it bothers some people, swap it out for "requires" and be done with it — hard to see how anyone can argue with that. --Trovatore (talk) 19:48, 8 September 2022 (UTC)
- @Kzqj You called my edit ungrammatical nonsense, you ignored my invitation to join the discussion and now you accuse me of being annoying? It's ridiculous.
- Several participants in our discussion tried to explain to you the difference between syntax and semantics but their efforts were in vain, and you still confuse the two terms. If you carefully read the article Colorless green ideas sleep furiously, you will have some idea. If you are a native English speaker then, by definition, you do not know much about English grammar - everything seems natural and automatic and works pretty well but - as you can see - not always. Just because you subconsciously use both syntax and semantics doesn't make you a linguist. 85.193.215.210 (talk) 02:26, 10 September 2022 (UTC)
- I think you're being unreasonably pedantic (even by my standards) and taking a very aggressive tack that is not likely to result in productive collaboration. That said, sure, it would be more complete to say that the diagnosis no longer requires the procedure, given the existence of modern equipment. I would suggest rewording along those lines, or perhaps combining with passive voice (say, The risky procedure is no longer needed given modern equipment), rather than trying to shoehorn in "necessitates". --Trovatore (talk) 01:44, 9 September 2022 (UTC)
- @Trovatore Your explanation was comprehensive and as clear as possible, and I admire your patience. Thanks :-) 85.193.215.210 (talk) 17:15, 9 September 2022 (UTC)
- Trovatore, have you looked at the history of the article Beryllium? Before this thread was started, Kzqj had already reworded the sentence in what seems an unobjectionable manner, without using needs, necessitates, or requires (although the edit summary was perhaps not the best choice). I'm really surprised that this thread has gone on so long, and I heartily concur with his or her observation that "arguing about whether this sentence is grammatically incorrect nonsense or just nonsense" is a waste of time. Deor (talk) 02:05, 9 September 2022 (UTC)
- It's a form of the pathetic fallacy, except that the equipment is not part of nature. Card Zero (talk) 17:16, 10 September 2022 (UTC)
- No it isn't. I don't know where this idea that "need" cannot be used in this way comes from, but it certainly isn't from speaking English. There's no suggestion of intent or anything like it DuncanHill (talk) 18:44, 10 September 2022 (UTC)
- The idea comes from scientific writing, as in Pathetic_fallacy#Science. Being hard-line about it is silly, I agree. Card Zero (talk) 02:24, 11 September 2022 (UTC)
- But "needs" simply doesn't qualify. It does not figurative, and it does not imply any human quality. DuncanHill (talk) 02:28, 11 September 2022 (UTC)
- Well, the word has several meanings, and it might be understood as "the equipment feels that it must have the procedure" or "the equipment is morally obliged to be involved in the procedure", by an obstinate person. Unlikely, yes. Card Zero (talk) 02:56, 11 September 2022 (UTC)
- But "needs" simply doesn't qualify. It does not figurative, and it does not imply any human quality. DuncanHill (talk) 02:28, 11 September 2022 (UTC)
- The idea comes from scientific writing, as in Pathetic_fallacy#Science. Being hard-line about it is silly, I agree. Card Zero (talk) 02:24, 11 September 2022 (UTC)
- No it isn't. I don't know where this idea that "need" cannot be used in this way comes from, but it certainly isn't from speaking English. There's no suggestion of intent or anything like it DuncanHill (talk) 18:44, 10 September 2022 (UTC)
- You could replace "no longer needs" with the word "obviates". ←Baseball Bugs What's up, Doc? carrots→ 02:34, 9 September 2022 (UTC)
- I propose we replace needs with aardvark, to generate:
- Modern diagnostic equipment no longer aardvark this highly risky procedure.
- This is ungrammatical!! Just in the interest of calming the waters and finding some common ground we all can agree on, you understand. Mathglot (talk) 07:11, 11 September 2022 (UTC)
- Finally, some ungrammatical nonsense. Card Zero (talk) 18:42, 11 September 2022 (UTC)
- I propose we replace needs with aardvark, to generate:
Universal Wikipedia
[edit]One might think this question is better suited to the computing desk, but I’m going to give it a go here: what is standing in the way of a universal Wikipedia, where everyone contributes in their native language to one single database, but their contributions appear instantaneously in the chosen language of the reader or editor? Currently, each language has their own separate Wikipedia instead of a universal database. How feasible is this from a language perspective rather than presenting it solely as an engineering problem? Viriditas (talk) 00:57, 8 September 2022 (UTC)
- See m:Abstract Wikipedia. Probably the most fundamental issue this project will face (aside from the actual software development) is gaining the trust and support of the community on a large scale. I don't think language would be the greatest barrier as long as the Wikidata model could be adapted successfully to Abstract. Shells-shells (talk) 02:06, 8 September 2022 (UTC)
- Thank you. I’m reading through the documentation now. I really hope they are successful. I obviously live in my own Private Idaho (pop. 1), as I have a hard time imagining how anyone could oppose such a wonderful, beautiful idea. Viriditas (talk) 07:44, 8 September 2022 (UTC)
- Part of the issue with that idea, as beautiful as it is, is that it presupposes some sort of platonic ideal language; where each individual language could be translated into and out of to allow for smooth, easy translation. Language is far too varied for that, there are linguistic features which defy translation into other languages; and I don't just mean vocabulary such as "single words in one language that don't have a single word in another language", I mean like entire features of syntax and semantics that don't apply in other languages. How does one translate from English into a language that has clusivity, for example, or evidentiality? These features are not really machine translatable, and still require the nuance of human translation to make work. --Jayron32 12:44, 8 September 2022 (UTC)
- I may not perfectly understand the goals of Abstract Wikipedia, but it seems to me that it is attempting to provide a skeleton of basic knowledge upon which meatier content can be applied through the traditional Wikipedia model. It would reduce the amount of redundant information added across different-language articles. I agree that it's probably unfeasible to provide universal prose translations of entire articles, but surely the barest facts can be expressed in a language-agnostic way. Statements like "Thing X had Attribute Y at Time Z" seem conceptually simple enough to be translated by machine, and a solid, well-sourced group of such statements would form a backbone of stability upon which more readily changeable content could be added. Shells-shells (talk) 17:12, 8 September 2022 (UTC)
- The one thing I am wary of with Abstract Wikipedia is that the Abstract Wikipedia "articles" (written by the volunteers in Wikidata/Wikifunctions code) are supposed to be CC BY-SA, but the WMF wants to keep the option open to make the human-language output CC0 (i.e. public domain). That's just completely crazy to me. Firstly, a human-language translation of an article – even if written in a meta-language – can never have a more permissive licence than the original, and secondly the "viral" nature of the CC licence which ensures that any derivatives are also CC BY-SA (and not proprietary) is lost. The project has apparently been making slow progress; Google recently seconded a number of staff to the Abstract Wikipedia project. My feeling is Google's interests are paramount in that project. Andreas JN466 00:57, 11 September 2022 (UTC)
- Part of the issue with that idea, as beautiful as it is, is that it presupposes some sort of platonic ideal language; where each individual language could be translated into and out of to allow for smooth, easy translation. Language is far too varied for that, there are linguistic features which defy translation into other languages; and I don't just mean vocabulary such as "single words in one language that don't have a single word in another language", I mean like entire features of syntax and semantics that don't apply in other languages. How does one translate from English into a language that has clusivity, for example, or evidentiality? These features are not really machine translatable, and still require the nuance of human translation to make work. --Jayron32 12:44, 8 September 2022 (UTC)
- Thank you. I’m reading through the documentation now. I really hope they are successful. I obviously live in my own Private Idaho (pop. 1), as I have a hard time imagining how anyone could oppose such a wonderful, beautiful idea. Viriditas (talk) 07:44, 8 September 2022 (UTC)
- As a smaller problem, Chinese Wikipedia has meta:automatic conversion between simplified and traditional Chinese. However, there are two Belarussian Wikipedias with different orthographies. The English Wikipedia tries to a single English dialect for each of its pages. --Error (talk) 12:18, 8 September 2022 (UTC)
- As a matter of comparison though, the differences among varieties of English are very slight; it's a surprisingly homogeneous language across the various parts of the world where it is spoken. The differences between dialects are rarely severe enough to cause major confusion. At most, such distinctions make a dialect marked rather than truly confusing (which is to say, it can present a bit of awkwardness without obscuring meaning), and even then, most native English speakers are aware enough of the minor dialectical differences (such as slight orthography distinctions like color/colour, or vocabulary differences like soccer/football) that having a single universal English Wikipedia causes no real problems. This is a greater issue in languages where the distinctions are stark enough to create significant understanding issues, for example there are multiple Serbo-Croatian language Wikipedias; despite the fact that the four main standard varieties of Serbo-Croatian are fully mutually intelligible, there are various both political and linguistic reasons (including using different alphabets) that make it more practical to have different such Wikipedias. The English language suffers neither from major political divisions (the Anglosphere being largely on friendly cultural and political terms for centuries) nor major linguistic differences that make separate Wikipedias necessary. It would only be necessary if, for example, the English variety in England were still written with futhorc while American English were written in the Latin script. We don't have those kinds of problems. --Jayron32 12:35, 8 September 2022 (UTC)
- My impression was that the Serbo-Croatian split was almost entirely political in nature. There might be better examples. (And Serbian could be written in both the Cyrillic and Latin alphabets, by the way.) 惑乱 Wakuran (talk) 21:48, 8 September 2022 (UTC)
- As a matter of comparison though, the differences among varieties of English are very slight; it's a surprisingly homogeneous language across the various parts of the world where it is spoken. The differences between dialects are rarely severe enough to cause major confusion. At most, such distinctions make a dialect marked rather than truly confusing (which is to say, it can present a bit of awkwardness without obscuring meaning), and even then, most native English speakers are aware enough of the minor dialectical differences (such as slight orthography distinctions like color/colour, or vocabulary differences like soccer/football) that having a single universal English Wikipedia causes no real problems. This is a greater issue in languages where the distinctions are stark enough to create significant understanding issues, for example there are multiple Serbo-Croatian language Wikipedias; despite the fact that the four main standard varieties of Serbo-Croatian are fully mutually intelligible, there are various both political and linguistic reasons (including using different alphabets) that make it more practical to have different such Wikipedias. The English language suffers neither from major political divisions (the Anglosphere being largely on friendly cultural and political terms for centuries) nor major linguistic differences that make separate Wikipedias necessary. It would only be necessary if, for example, the English variety in England were still written with futhorc while American English were written in the Latin script. We don't have those kinds of problems. --Jayron32 12:35, 8 September 2022 (UTC)
- Machine translation will never be perfect but it has come along in leaps and bounds. Try playing around with DeepL.com – paste the lead of a random German or Czech or Spanish or Japanese Wikipedia article in and see what you get. Andreas JN466 07:02, 9 September 2022 (UTC)
- I tried machine-translating the opening paragraph of Cat from English to French and then back to English to demonstrate the loss of information that would result.
- Instead of being "referred to as the domestic cat" it was now "called the domestic cat", which makes it sound more like a personal name or nickname.
- "The family Felidae" became "the Felidae family", and "the wild members of the family" became "the wild family members," making it sound like a specific pair of individual cats and their offspring, sharing the surname Felidae.
- Feral cats were now said to "vary freely" rather than "range freely".
- "Various cat registries" became "various registers of cats", replacing the idiomatic phrase with a more abstract idea of anything that registers a cat. Such as an AI, perhaps, or a handheld clicker for counting cats with?
- Only the third of these deviations from the original is really serious, but they all make the text less expressive, and more vague. Card Zero (talk) 14:40, 10 September 2022 (UTC)
- Card_Zero -- The most classic case of round-trip translation supposedly occurred as part of early 1960s machine translation research, when the Bible quote "The spirit is willing, but the flesh is weak" was translated from English to Russian back to English, resulting in "The vodka is good, but the meat is rotten" (mentioned on article Literal translation)... -- AnonMoos (talk) 19:42, 10 September 2022 (UTC)
- @Card Zero: Which translation software did you use? Neither DeepL nor Google reproduce your results for me. For reference:
- Original
- The cat (Felis catus) is a domestic species of small carnivorous mammal. It is the only domesticated species in the family Felidae and is often referred to as the domestic cat to distinguish it from the wild members of the family. A cat can either be a house cat, a farm cat, or a feral cat; the latter ranges freely and avoids human contact. Domestic cats are valued by humans for companionship and their ability to kill rodents. About 60 cat breeds are recognized by various cat registries.
- DeepL translation to French and back
- The cat (Felis catus) is a domestic species of small carnivorous mammal. It is the only domesticated species in the felidae family and is often referred to as a house cat to distinguish it from the feral members of the family. A cat can be a house cat, a farm cat or a feral cat; the latter moves freely and avoids contact with humans. Domestic cats are valued by humans for their companionship and ability to kill rodents. Approximately 60 breeds of cats are recognized by various cat registries.
- Google translation to French and back
- The cat (Felis catus) is a domestic species of small carnivorous mammal. It is the only domesticated species of the Felidae family and is often called the domestic cat to distinguish it from wild members of the family. A cat can be either a domestic cat, or a farm cat, or a wild cat; the latter moves freely and avoids human contact. Domestic cats are valued by humans for their companionship and their ability to kill rodents. About 60 cat breeds are recognized by various cat registries.
- Andreas JN466 00:51, 11 September 2022 (UTC)
- Well, an inferior one, obviously (I used Lingvanex, because my browser has a handy button for it). I tried the next paragraph in DeepL, and there were errors. Grunting was replaced with snarling, and the phrase A predator that is most active at dawn and dusk (crepuscular) became A very active predator at dawn and dusk (twilight), so that it sounds like less of a scientific fact and more of an anecdotal observation. Then sounds too faint or too high in frequency became sounds that are too low or too high in frequency, changing the first concept into a different one. So there's a bit of information decay. It seems to be making some reasonable contextual assumptions, but the problem is that translators need to make assumptions. If they're humans, they can debate among themselves about the veracity of the results. There's not, as Jayron said, a Platonic ideal of the meaning of the paragraph, which could be extracted, stored in a bottle and released into an arbitrary language. The words in the text exist in relation to our minds and culture, and we have to tend to them continually and herd them around to keep them in line with what they're supposed to say as we understand it. Card Zero (talk) 02:00, 11 September 2022 (UTC)
- Well, you have to allow for the fact that you are using the process twice: which is not a realistic use case. Secondly, if you ask two human translators to do the same thing – ask one to translate into French, and another to translate it back – you will also get decay, and perhaps different types of errors (a human sometimes forgets to translate a word, and MT generally does not ... although, oddly enough, DeepL missed out a whole sentence when I first used it on the Cat paragraph, something which happens very rarely in my experience, and I use it a lot). Andreas JN466 09:31, 11 September 2022 (UTC)
- Of course this also applies to human translators, I almost mentioned that myself. The point is that there is decay, which then needs actual intelligent attention from humans with a broad understanding of the subject and related subjects and the wider world. They need to be aware of what phrases sound like they mean, what they imply, and which details of the subject are crucially important. They need to be alert to common misconceptions and likely errors, and things that don't add up. Us editors are not supposed to do original research and write what we think is true, but on a low level it's kind of crucial that we do. Translate an article, and it needs all this intelligent scrutiny afterward. There's something intriguing to say here about whether its meaning is "the same" after translation. Clearly what we're trying to do by wielding our mighty human intelligence is to maintain the same meaning, and AIs can't because they can't judge intended meaning or say what matters. However, all the words in the original language had the potential to be misunderstood in multiple ways. ("Faint" as a synonym for "low", for instance.) if your take on the intended meaning is different - if you have a bizarre idea about what was intended - then the article will seem to be mistranslated. Card Zero (talk) 18:31, 11 September 2022 (UTC)
- The same sorts of issues are likely to arise in Abstract Wikipedia as well, aren't they, where the entire process of translation from the metalanguage to the human (African, Indian etc.) language is supposed to be automated as well.
- I wonder what the best starting point for translation is. Abstract Wikipedia aims to create a representation of Wikidata items and their attributes or relationships in an abstract computer language, and to then translate the articles written in that computer language into whatever human languages are required (i.e. whenever there isn't a Wikipedia article on the topic in the relevant language).
- What I'm wondering is whether that is more likely to achieve success in the fastest possible time than starting with articles written in (grammatically) Simple English, and then throwing existing machine translation methods at that. Andreas JN466 22:45, 11 September 2022 (UTC)
- Of course this also applies to human translators, I almost mentioned that myself. The point is that there is decay, which then needs actual intelligent attention from humans with a broad understanding of the subject and related subjects and the wider world. They need to be aware of what phrases sound like they mean, what they imply, and which details of the subject are crucially important. They need to be alert to common misconceptions and likely errors, and things that don't add up. Us editors are not supposed to do original research and write what we think is true, but on a low level it's kind of crucial that we do. Translate an article, and it needs all this intelligent scrutiny afterward. There's something intriguing to say here about whether its meaning is "the same" after translation. Clearly what we're trying to do by wielding our mighty human intelligence is to maintain the same meaning, and AIs can't because they can't judge intended meaning or say what matters. However, all the words in the original language had the potential to be misunderstood in multiple ways. ("Faint" as a synonym for "low", for instance.) if your take on the intended meaning is different - if you have a bizarre idea about what was intended - then the article will seem to be mistranslated. Card Zero (talk) 18:31, 11 September 2022 (UTC)
- Well, you have to allow for the fact that you are using the process twice: which is not a realistic use case. Secondly, if you ask two human translators to do the same thing – ask one to translate into French, and another to translate it back – you will also get decay, and perhaps different types of errors (a human sometimes forgets to translate a word, and MT generally does not ... although, oddly enough, DeepL missed out a whole sentence when I first used it on the Cat paragraph, something which happens very rarely in my experience, and I use it a lot). Andreas JN466 09:31, 11 September 2022 (UTC)
- Well, an inferior one, obviously (I used Lingvanex, because my browser has a handy button for it). I tried the next paragraph in DeepL, and there were errors. Grunting was replaced with snarling, and the phrase A predator that is most active at dawn and dusk (crepuscular) became A very active predator at dawn and dusk (twilight), so that it sounds like less of a scientific fact and more of an anecdotal observation. Then sounds too faint or too high in frequency became sounds that are too low or too high in frequency, changing the first concept into a different one. So there's a bit of information decay. It seems to be making some reasonable contextual assumptions, but the problem is that translators need to make assumptions. If they're humans, they can debate among themselves about the veracity of the results. There's not, as Jayron said, a Platonic ideal of the meaning of the paragraph, which could be extracted, stored in a bottle and released into an arbitrary language. The words in the text exist in relation to our minds and culture, and we have to tend to them continually and herd them around to keep them in line with what they're supposed to say as we understand it. Card Zero (talk) 02:00, 11 September 2022 (UTC)
- I tried machine-translating the opening paragraph of Cat from English to French and then back to English to demonstrate the loss of information that would result.
- While I love a good translation discussion as much as anybody, I think the major objection to this idea would not be about translation at all, but something much more human. In my case, for example, I would oppose any article appearing automatically in English that was edited by users of a language whose Wikipedia culture and local policies and guidelines were weaker than ours or observed less than ours at en-wiki. I've participated at a few, and some are appalling. Some Wikipedia language cultures simply don't observe the guidelines as much, and consider them more of a "nice-to-have" than an obligation. I think the different levels of adherence to verifiability, NPOV, and other important guidelines would be the downfall of a universal project, even with a perfect translator available. Mathglot (talk) 07:28, 11 September 2022 (UTC)
- Surely this is an argument in favor of greater universalism, no? If different wikis have differing policies, wouldn't a universal wiki help to bring those policies closer to a single standard? Different Wikipedias are indeed, arguably, too different from one another than they should be, but a universal wiki would presumably help to reduce that difference. It would certainly help to mitigate some of the serious problems that cultural isolation causes within wikis.The argumentum ad naturam humanam has not really been convincing to me ever since the Wikimedia movement came to exist. The success of this very site is directly contrary to what a common-sense understanding of human nature would tell us; people and communities can be a lot more malleable than we typically give them credit for. That said, I agree in principle that translation is not the issue. In essence, we are both restating the same idea: Abstract Wikipedia will not be successful unless contributors trust it, and gaining that trust will probably be difficult. However, the existence of Wikipedia itself (let alone Abstract Wikipedia) proves that such trust can be gained to kickstart—almost bootstrap—such a community into existence. Shells-shells (talk) 08:00, 11 September 2022 (UTC)
Is there perhaps a legal problem here? Ultimately, Wikipedia contributors can (at least in theory) be held legally accountable for things they write. Who would be responsible for say libel generated by an error in machine translation? I can't imagine that software developers would accept responsibility, and I can't imagine many contributors would want to take the risk of being sued for something they wrote being mistranslated into a language they can't understand. AndyTheGrump (talk) 23:02, 11 September 2022 (UTC)
If you found an error in an article, how would you correct it, if it was written in a language you didn't understand? — kwami (talk) 23:40, 11 September 2022 (UTC)
A trains
[edit]I recently wrote this in the article Espoo: "A trains travel along the Leppävaara city railway with a terminus at Leppävaara."
Now when it says "A trains", it means "trains on the line A". It is not a typo for "trains" or "a train". Is there danger of confusion or is it fine as it is? JIP | Talk 01:03, 8 September 2022 (UTC)
- You could clarify by making the A stand out in some way, such as by putting quotes around it, italicizing it, or hyphenating as A-trains. ←Baseball Bugs What's up, Doc? carrots→ 02:40, 8 September 2022 (UTC)
- The 1st option is grammatically fine, but doesn't scan well. Go with the 2nd option ("Trains on the A line travel along...") -- problem solved. 136.56.52.157 (talk) 02:49, 8 September 2022 (UTC)
- Note that two sentences before, it says "Only A and L trains stop at Mäkkylä", so the confusion is not all that great. Still, the second phrasing is clearer. Maybe we should ask Duke Ellington for his opinion? Clarityfiend (talk) 02:57, 8 September 2022 (UTC)
- I can't prove this, but someone once told me that the Lawrence Welk group performed that song on the TV show, and Welk introduced it as "Take A Train". — Preceding unsigned comment added by Baseball Bugs (talk • contribs) 05:28, 8 September 2022 (UTC)
- Note that two sentences before, it says "Only A and L trains stop at Mäkkylä", so the confusion is not all that great. Still, the second phrasing is clearer. Maybe we should ask Duke Ellington for his opinion? Clarityfiend (talk) 02:57, 8 September 2022 (UTC)
- You could use "The A trains travel along ...", making it abundantly clear that the "A" is not the English indefinite article capitalized in sentence case. --Lambiam 08:22, 8 September 2022 (UTC)
- A trains is rather odd-looking. A-line trains, or "A" trains, or A-trains all feel better to me. DuncanHill (talk) 09:55, 8 September 2022 (UTC)
- You can check how this is done in other A Train articles, for example A (New York City Subway service) (first thing I think of when you say "A Train" other than User:A Train): "On December 11, 1988, A trains began running ..." —Kusma (talk) 10:01, 8 September 2022 (UTC)
- See the typography of the song title "Take the "A" Train" for an option. --Jayron32 12:02, 8 September 2022 (UTC)