|Given the guidelines currently in place, why are we ADDING "adverb" to prepositional phrases?|
Sorry to bother, but the past tendency to mark prepositional phrases "adverb" is contrary to current guidelines, yet the label is being added to new entries:
The guidelines include the following:
"OTHER WORD CLASSES
Whole sentences, phrases (including prepositional constructions with a noun or pronoun object), examples and fixed expressions, as well as cardinal numbers, do not belong to any word class and should be marked with "[none]"."
Given this, we shouldn't be adding "adverb" to such entries, should we? I wouldn't have posted this, except that quite a number of people seem to have voted contrary to the guidelines here.
|Indeed, they did vote against the guidelines.||#519789|
Wouldn't be the first time, I suppose...
I used to vote against this guideline myself, but I am trying to reform now :-) (It's not doing much good, though!)
Another case where a vote was cast without acknowledging concerns raised in earlier comments - see the contrib. discussion here: http://forum.dict.cc/?pagenum=11089&hilite=519690&kw=#entry...
Some contributors have a very high final/accepted vote ratio, all without comment. Hmm.
|maybe one could install a "hold" option||#519795|
I don't know whether this might work and it would probably put a huge workload on Paul to install it.
If one could put a hold on entries that one thinks need to be changed it would keep people from "waving it through" without regard to objections raised by native speakers. I know that would further slow an already lagging process. There is also a good chance it would be abused by fun-loving hackers. (One should not offer this option to non-registered voters)
Another idea would be to require voters to check a box next to comments left by other voters with "agree" or "disagree." That would still not stop industrious people from voting on chance of success rather than merit, but it would at least guarantee that comments get looked at.
While we are at it, I'm not happy with what I presume are the results of that "automatic suggestion" window. I see lots of unregistered voters and newbies voting for original entries and only rarely ones that have actually been checked. This way I've seen things that already had 4 or 6 votes and still were very wrong. (Whether in format or content.)
It seems to me that the designation "adv" should include adverbial phrases. (of time, manner, place)
This would be helpful to clients. It would also solve the problem when one language uses a single word that is clearly an adverb, while the other uses two or more words, and if this tag is limited strictly to a single word, it should not be so marked.
(e.g. a simple example: later = plus tard.) One is an adverb (single word); the other is an adverbial phrase of time. The sensible thing would be to extend the meaning of adv to include adverbial phrases.
The definition of [adv] in the guidelines is not only adverb but also adverbial. An adverbial by definition includes adverbial phrases (see, for example, http://www.answers.com/topic/adverbial). Hence, the listing or submission of phrases as [adv] ought not to be a problem. Tagging prepositional phrases as [adv] is quite another matter, and such erroneous listing should be picked up by the checkers. Big voters (with power 5) and unreflective voters (with any level of power) will always skew things in the checking process. I myself am often guilty of not weighing up all the possible qualifications that could be made to an entry. Whether a so-called "hold option" would improve things is open to further debate, but I suspect good old "human error" would still win through.
|However this may be -||#519839|
I'm hopeless with anything but the most basic grammar (terms) - this was cleary a mistake on my side, not just in the sense that adv. might not be correct here, but also in the sense that I reopened the entry quite unintentionally. Probably I wasn't aware that I had cast the final vote. I'd never normally reopen a verified entry for 'grammar' reasons, unless for instance a translation of nouns was hanging around at some odd place, because the 'noun'-tag was missing.
If 'in a moral sense' were an adverb or adverbial phrase, I think it would be correct according to the guidelines to tag it adv. The phrase it not that long, and I thought the main purpose of all that grammar tagging was putting translations into the proper category.
Anyway, sorry for the mistake.
|Gott sei Dank, ich versteh nix von Grammatik!||#519842|
Deshalb fühle ich mich auch so FREI von Zwängen. Und vote trotzdem!
|on the other hand:||#519848|
im übertragenen Sinn - figuratively [adv.]
im gegenläufigen Sinn - reversly [adv.]
Where is the difference? [For practical purposes?]
|Thanks Lisa4dict for the technical suggestions!||#519854|
I thought about them, but unfortunately I don't think the benefits would outweigh the additional problems that would arise. I'm always open to suggestions, but the system is already even a bit too complex. If there would be a way to solve problems by removing complexity, that would be great. The whole system should be as easily understandable as possible.
|Adverbial and non-adverbial phrases||#519857|
My previous post had to be made somewhat in haste, as I had some urgent business. Having looked at the matter further, and at the example cited, I would like to add this.
The example cited is not even an adverbial phrase and should indeed be tagged "none". (In rather broad terms: it does not say "when" or "where" or "why", and it is arguable whether it says "how".
Repeating, for convenience, the GL mentioned above:
OTHER WORD CLASSES
Whole sentences, phrases (including prepositional constructions with a noun or pronoun object), examples and fixed expressions, as well as cardinal numbers, do not belong to any word class and should be marked with "[none]".
Comment: A true adverbial phrase such as "plus tard" is not a prepositional construction, nor does it have a noun or pronoun object, therefore the GL does not apply to it, or to similar constructions.
The entry quoted (in a moral sense = im moralischen Sinne) has a preposition and a noun object, so this Guideline does apply; it is not an adverb, nor a true adverbial phrase. I think we are arguing about different concepts.
Können wir das nicht etwas vereinfachen?
Wie kann ich erreichen, dass eine (kurze und häufige) Wendung an auffindbarer Position steht? Darum geht es doch eigentlich, oder?
Dinge ohne word classification landen in einem unübersichtlichen Haufen, indem man sie dann u. U. suchen muss, wie die sprichwörtliche Stecknadel im Heuhaufen. (Der - falsche - adv.-tag hilft da übrigens auch nicht. Oh, Jammer! Wenn ich Sinne eingebe, geht's noch, aber wenn ich moralisch eingebe, komm ich gar nicht hin.)
|Agree with Catesse and jb||#519890|
The thing is that "adv" and "prepositional phrase" are not mutually exclusive. My understanding of the rules is that whenever a "prepositional phrase" functions as an adverbial clause, i.e. modifies a verb, adjective or a complete sentence describing a state or an action, it overrides the "no word class" rules. I suppose that part of the confusion arises from the term "word class", which implies that it refers strictly to single words, when "part of speech" or "lexical category" would be a more fitting name. Overall, we should be looking more at an entry's function in a sentence rather than at morphological features. (That's also the reason why we got rid of the distinction between adjectives and "attributes", a constant cause of confusion and unnecessary quarrels.)
I also agree with Catesse: As this is a bilingual dictionary, it's often impossible to have single-word entries on both sides, because one language may denote a concept with a single words, while the other may need more than one word. Catesse mentioned the obvious "plus tard", which may not be the best example in this respect, as the two words stem from the fact that the comparative and superlative are formed differently in French and English. However, take a German adverb like "mittelmäßig", which might be translated as "in a mediocre manner/way/fashion"—is it incorrect to tag this entry as an adverb because there is a prepositional phrase on one side?
There is yet another problem: What is a "word"? Different languages form compound words in different ways. To use an example from Polish, "to rise from the dead" is zmartwychwstawać, which is a compound of z martwych wstawać – a literal equivalent of the English phrase in all but the missing spaces. Would it be "wrong" to tag this entry as a verb just because there is a phrase on the English side instead of a "single word" like on the Polish side?
On the whole, I see no reason why we should quit tagging adverbial clauses as "adv."
|need for a ruling||#519907|
The discussion till now shows that good arguments can be made for both positions -- with the result that after hearing out the discussion Paul will ultimately have to rule on the matter, though understandably he does not like having to make such rulings.
Much of the problem here, as in other cases, lies in the fact that the word class automatically applies to both sides of a translation pair, though in fact it may fit only the one or the other side, e.g. Thorsten's case of the Polish verb "to rise from the dead". Unless Paul can find a remedy for this problem -- personally I can see no practicable alternative -- we are left with three choices:
a) assign "none" whenever the same word class does not strictly fit both sides of a pair
b) assign a word class whenever that class strictly fits at least one side of the pair
c) assign a word class not on the basis of formal structure but rather on the basis of semantic function
According to the first alternative, "adv." would not be assigned, when a prepositional phrase stands on one side.
According to the second alternative, it would be assigned if on at least one side a simple adverb stands.
According to the third alternative, a prepositional phrase would count as an adverb if it answers questions of when, where, how, or by.
As ddr points out, the first alternative would exclude a large number of existing entries from the adverb class, burying them in the miscellaneous "none" class.
The second alternative would simply continue current, but not altogether consistent practice.
The third alternative would probably be difficult to apply for many contributors and voters and if enacted would entail re-opening many entries if consisent application should be demanded.
Personally, I tend toward the second alternative, but i will go along with whatever method is chosen
|tomaquinaten et al.||#519914|
Once upon a time, pupils were expected to know what an adverb was by the time they were about eleven. Or else!! Then it was decided that grammar was a waste of time and too hard for the little darlings.
Here, there has been a move back to basics, with threats that schools which do not teach basic literacy (including grammar) and numeracy (including multiplication drills) will lose Federal funding. (Most of the money for education comes from the State treasury. The central Govt just tops it up for extras and special needs.)
It does not take a genius to learn "when, how, where and why", although how to apply it on dict. in complex cases is a different matter. What most researchers would need to know is whether words that are similar are adverbs or adjectives. The rest is window dressing.
|From my point of view||#519917|
it doesn't make a difference for 99.9% of dict.cc's users. They look up the phrase to see the translation. They have the German term and want to know the equivalent in English, or vice versa.
If the dict.cc community is not able to decide whether this is an adverbial phrase or a prepositional phrase or even none or both, it will not matter for the user. The difference doesn't produce outvotes and it doesn't change the position of the entry in the search results. So, to be honest, if I wouldn't have been asked, I wouldn't care. As a contributor I wouldn't reopen an entry like this to add or remove the adv class.
The term "word class" is meant as "part of speech" or "lexical category". If the German adverb "mittelmäßig" would be added as "in a mediocre manner" I would still consider it an adverb (or an adverbial phrase). I guess this translates into tomaquinaten's alternative b), but I'm not sure if there even is a difference between b) and c) in practice. The clearest rule would be b).
with Paul, agree with Toms b suggestion, agree with Catesses 'window dressing' judgement.
My (retrospective) hope, that by adding a word class the thing would appear in the adjectives and adverbs section has turned out a sham. It's the number of words that counts.
For sombeody trained to use print dicts., dict.cc does have a few flaws. :))
Catesse: German speakers no doubt learn less about adverbs or Umstandswörter, as they were called once upon a time in my primary school days, they come across them, when they start to learn English, and it's normally pretty difficult for them to grasp the concept, but that the concept is that complicated, I learned only here.
I agree with your common-sense approach to this. However, there is a contradiction in the guidelines: On the one hand, they instruct us to tag both (single-word) "adverbs" and (multi-word) "adverbials" as adv. Which makes sense, too—in fact, anything else would be fairly nonsensical. Yet, another guideline instructs us that "prepositional phrases" mustn't be be assigned any part of speech at all, even though the vast majority of adverbials include a preposition. This is bound to generate trouble.
As for Catesse's argument—if we were to add "adv" only if there is a potential for confusion between adverbs and adjectives, we might as well throw the whole part-of-speech tagging system over board.
|Okay, so we'll have to change this section in the guidelines to make this clearer.||#520153|
Unfortunately I need to get going now.
Wording suggestions welcome!
What you thought I said was not what I meant to imply. I think that we should tag both single adverbs and adverbial phrases as adv., at least if the concept is a single word in one or other of the languages.
However, it is my belief that the majority of clients do not give two hoots about it; in fact, it is likely that they do not even notice the tags. People engaged in high-level translating are likely to have other sources than dict. at their disposal. They would probably use these for most of their work, and go to Forum for really difficult bits. Therefore the main dict. is likely to be of benefit largely to non-professionals and rank amateurs. This is only an estimate; there is no way of consulting users to ascertain their motives and to learn what they hope to find in dict.
Personally, I would seldom use dict. in translating a German or French text, but I would make lively use of it for Spanish or Dutch or Swedish, at least when their vocabulary is larger.
|Wording suggestion - get rid of contradictory wording in Guideline #4||#520631|
The rule that tells us to use "adv" for both single-words adverbs and "adverbials" (phrases) makes perfect sense, not least because very often a single-word adverb in one language needs more than one word in the other language: Just think zuhause vs. at home, nach Hause vs. home, morgens vs. in the morning. Many, if not most, adverbial phrases include a preposition (like "at", "nach", or "in") and a noun, (like "home" or "morning") in at least one of any two languages. That's why it just doesn't make any sense to have a rule like "... prepositional constructions with a noun ... do not belong to any word class and should be marked with "[none]". Instead of trying to enumerate what "doesn't belong to any word class", I would simply have the rule say that "Entries that do not belong to any specific word class (part of speech, lexical category), such as whole sentences, phrases, examples and fixed expressions, as well as cardinal numbers, should be marked with "[none]".
4;Catesse: I'm glad we agree after all. :) However, I still don't quite agree with your reasoning. True, there may be users who don't need those tags because they don't know what the difference between an adjective and an adverb is, just like there may be users who don't need them because they can tell the difference without them. However, just like any other dictionary, dict.cc has been tagging entries for their grammatical categories from the start, and if we do it, we may as well do it properly and consistently, regardless of the question "who needs" it.
I also think it's pretty futile to try and "profile" dict.cc users. I for one do have lots of "other sources than dict.cc at my disposal". Yet, dict.cc is often the first place I go when looking for translations. Not only because it's a lot quicker than looking up stuff in paper dictionaries, but also because it contains many specialist translations that I couldn't find in any of my own dictionaries or translation memories.
I would also suggest to allow experienced users, that can be trusted to understand word classes, to add or correct them without having to wait for the approval of everyone—as 99% of the time, such changes will be totally uncontroversial.
|Guideline 9 updated.||#520717|
Thanks for the suggestion, Thorsten! Muhamed and I both agree and we now use your wording.
Regarding word class changes: I originally intended to do just that, but then some controversy started about a specific VP5 contributor's class changes. That was even before the class was a field on its own. So I didn't use the already written code. As for today, I'm still uncertain if this may lead to errors, so I rather not change it for the time being.
+"we now use your wording"+: Thanks, it's an honor. ;)
As for the word class changes, I realize that controversies may erupt about anything, word classes being no exception (as this thread demonstrates). However, I still think that word class (or parts of speech) has less potential for conflict than other areas.
I'm thinking of situations like this one: http://contribute.dict.cc/?action=show-history&id=979797
I forgot to add to the word class. Along comes Bella, who takes care of this, but alas, adds one too many. As a result, the entry now has to go through the whole tedious verification process again.
Likewise, the entry that triggered this thread ("in a moral sense") should have been tagged "adv" per the spirit, if not the letter, of the guidelines (by now, also per their letter). Regardless, the tag was ultimately removed. I suppose a lot of those who voted to have it removed did so because they were annoyed to be bothered by a reopen at all—not because they subscribe to a strictly literal interpretation of one of two contradictory guidelines.
As a compromise, wouldn't it be technically possible to allow word-class changes without triggering the full reopen process at first? The involved users could simply be notified, and a full reopen would start only if someone objected to the change.
I have just had a look at the GL as they are on the DE-EN site at the moment.
As far as I can see, the adv tag can still be entered with a simple click. Did I misunderstand Paul comment?
The GL still do not address the question of how to handle entries where in one language there is a simple, one-word adverb, but the other language expresses the identical concept in two of three words.
I feel a bit like Alice in Wonderland at the Madhatter's Tea Party.
I'm not sure if I fully understand your suggestion. By not triggering the full reopen process, are you saying that the three-day grace period should not be triggered? This would mean that two VP5 contributors could change the word class and close the entry. Other previous voters wouldn't have the chance to agree, so they would be outvoted. I don't think this would work.
The guidelines at the top right describe the word classes. For "adv" the explanation "adverb/adverbial" is given, which means that both adverbs and adverbial phrases should be tagged as "adv".
What I meant is this: Experienced users should be able to add missing or change incorrect word-class tags with immediate effect, and without affecting anyone's rate of accepted votes. The original author and users who voted on the entry would receive a message telling them that the word class has been changed. If they agree (as most of the time they will), they can simply do nothing, without incurring an outvote. If they disagree, they may object—only in this case would there be a full re-verification process, requiring the approval of users with a certain VP between them. (Similar to the reopens by VP1 users, that can now be stopped without a full discussion.)
Reverting reopens is a bit different, because no outvotes occur if the entry is set back to the state it was before.
But with your suggestion, the previous votes would basically have to be changed by the system to avoid outvotes. This would mean that it wouldn't be necessary to check the word class anymore when voting. If it would be wrong and the entry would be reopened, there would be no outvote, so there wouldn't be much of a reason for voters to take care. This raises the question: Should the word class be in fact removed from the voting accuracy scheme?
Well, I have no idea how difficult it would be to implement such a change technically. If it would require the word class to be removed from the voting accuracy altogether, this would be a more fundamental change than I envisaged. However, it may be worthwhile to consider this: I can say from my own experience that the word class is easily forgotten, because it's added automatically most of the time. Also, it's easy to make a wrong click. In either case, adding a wrong class or none at all is no fundamental mistake (such as a misspelling or a wrong gender). Also, quite often you see entries hang on forever because votes are distributed more or less evenly across versions with and without word class tags. Removing the word class from the accuracy calculation might also redress this, as people can vote for the complete version without having to worry about causing poorly justified outvotes.
"The GL still do not address the question of how to handle entries where in one language there is a simple, one-word adverb, but the other language expresses the identical concept in two of three words. "
I see no reason why this case should be treated any differently than cases where there are single words on both sides, or multiple words on both sides. Why on Earth should an entry like "at home / zuhause" be tagged differently than "at home / zu Hause"? In terms of function, one-word adverbs are just a subset of the superordinate class of adverbials. Whether things are spelled in one, two or more words, is actually just a matter of orthographic convention ("zu Hause" vs. "zuhause", "zmartwychwstać" vs. "z martwych wstać"), and has nothing to do with grammatical part-of-speech classification.
Yes, I think that from a technical point of view your suggestion would have the same effect as removing the word class from the outvote system altogether, which would make quite a big difference. I originally intended to treat word classes like subjects (which can be changed without requiring confirmation), but then didn't, because I think that this could cause more harm than an incorrectly assigned subject.
I guess changing this would require a whole new discussion.
Yes, I see the adverb/adverbial at the top of the Abbreviations section. Fine.
But I was still reading - or thought I was reading - in the GL section that phrases beginning with a preposition should be given the tag [none]. This seemed to be in conflict with the Abbreviations. This instruction is no longer there, so that is fine too.
|Ist es jetzt ADVERB oder [none]||#522703|
Tut mir leid, aber ich verstehe es immer noch nicht!
Auch wenn ich obigen Pfad hin und her, vor und zurück lese und sogar mit ihm ins Bett geh.
Vielleicht liest ja jemand von den Experten meine Frage und gibt darauf eine Antwort.
JA, ist richtig verifiziert - NEIN, ist nicht richtig verifiziert.
As it's an adverbial, it should have been tagged "adv" from the start. However, there was another rule that prepositional phrases (which most adverbs are) should be tagged "none". When this was tagged to "adv", the voters decided to violate rule #2 and apply rule #1. Then there was a backlash—it was decided to apply rule #2 and violate #1 instead. Now that the contradiction has been removed, it should definitely be tagged "adv".
|Danke Thorsten! Ich korrigiere dann den Eintrag, ...||#522715|
weil ich ihn als REFERENZ brauche. Muß ja immer wieder nachschauen.
Sorry, I completely overlooked your statement from a few days ago:
"I originally intended to treat word classes like subjects (which can be changed without requiring confirmation), but then didn't, because I think that this could cause more harm than an incorrectly assigned subject. "
Maybe, maybe not. In fact, incorrect subjects, or AE/BE tags for that matter, can cause quite a lot of confusion. E.g., "Lab(o)ur" means work in general usage, but "Wehen" in medical usage, etc. "Capital" can be a city (general usage) or money (financial usage), etc. I think we can all come up with countless examples of words that mean completely different things in different areas. That's why I'm not sure that a wrong word class can do so much more harm than a wrong subject —the opposite case could be made just as well. But you're right, this is another topic and shouldn't be dealt with in this thread.
Please log in to post an answer to this thread - or post a new question.
|nach oben | home||© 2002 - 2019 Paul Hemetsberger | Impressum / Datenschutz|
Dieses Deutsch-Englisch-Wörterbuch basiert auf der Idee der freien Weitergabe von Wissen. Mehr Informationen!
Enthält Übersetzungen von der TU Chemnitz sowie aus Mr Honey's Business Dictionary (Englisch/Deutsch). Vielen Dank dafür!
Links auf dieses Wörterbuch oder einzelne Übersetzungen sind herzlich willkommen! Fragen und Antworten
Enthält Übersetzungen von der TU Chemnitz sowie aus Mr Honey's Business Dictionary (Englisch/Deutsch). Vielen Dank dafür!
Links auf dieses Wörterbuch oder einzelne Übersetzungen sind herzlich willkommen! Fragen und Antworten