Hunting homographs

I've added a feature to deal with homographs—words like "read" that correspond to two or more non-English words.

Two examples:
1. This page (when a message has "read")
2. http://www.librarything.com/groupzeitgeist.php?group=hogwartsexpress

Both appear normal on the English version. If you're in the French version, they appear as "read" until you try to translate them. Then you see:

read [Note: Imperative]
read [Note: adjective]

So, two down. How many to go?

Send me your homographs, I'll give them this treatment.

t

This maybe doesn't belong to these, but not in every language the abbreviation of May is not the same as the whole name of the month:
May in Finnish is toukokuu (abbreviation tou),
in Italian maggio (abbreviation magg.)

Are the month abbreviations a problem now. Shoot me an example?

>3 timspalding:
Well, it is not among the biggest problems in translations...

If I translate May to toukokuu, it is then toukokuu also there where abbreviations are used. I cannot remember where it is, but I could look for it later. But as I said, it is not very important. Inflectional forms are the ones that make me desparate.

Okay, give me examples of the inflectional forms. I can change them now.

They are everywhere, but here is one example:

Zeitgeist page
in English:
overview | books | authors | tags | users | by language | helpers

In Finnish:
http://fi.librarything.com/zeitgeist

yleiskatsaus | kirjat | tekijää | avainsanat | käyttäjää | kieliryhmittäin | avustajat

The word 'authors' is translated to 'tekijää', but it should be 'tekijät'.
If I change the translation to 'tekijät', it is incorrect in other places, where there should be (for example) 10 tekijää (= 10 authors). Same problem with the word 'käyttäjää' (= users).

More about inflectional forms in Finnish:
http://www.finland.fi/netcomm/news/showarticle.asp?intNWSAID=25835

If it is possible to solve this problem, I can send more examples.

Well, I'm going to need specifics, and how to label them. What confuses me here is that both situations you describe seem like the nominative. At least they would be so in Latin. I understand that, if the "authors" is preceded by "SEE," it would throw authors into the accusative.

I am afraid that this will be terribly complicated...

The words
books | authors | tags | users | helpers
in Zeitgeist page are meant to be plural nominative forms, aren't they?

These are the Finnish nominative forms in plural: kirjat, tekijät, käyttäjät.

These are partitive forms
in singular: kirjaa, tekijää, käyttäjää
in plural: kirjoja, tekijöitä, käyttäjiä.

The partitive forms are used with numerals like this (partitive noun is in singular!):
10 kirjaa (= 10 books), 10 tekijää (= 10 authors), 10 käyttäjää (10 users). BUT: 1 kirja, 1 tekijä, 1 käyttäjä.

If I change this 'tekijää' to 'tekijät' it's OK on this page but not everywhere.

Does this clarify the problem at all? I am not very good at explaining...

Jukka Korpela has made a table of cases in Finnish and their corresponding English prepositions:
http://www.cs.tut.fi/~jkorpela/finnish-cases.html

So, with a numeral Finnish doesn't use the plural, but a partative singular, like "ten OF books"?

Can I suggest that it might be harder to change LibraryThing for this than to change Finnish itself? ;)

I think the solution would be to stop doing phrase->phrase for select languages, pages or passages. That might be harder. I'll look into it.

>10 timspalding:

Thanks, Tim, for listening to my woes.
I have gotten the impression that in many languages it would be much easier to translate whole sentences or passages than words or phrases. So it would be very helpful if you could change the translation system for us.

No, I mean changing from phrase->phrase into phrase+page->phrase. Sorry, is that unclear?

>12 timspalding:

Oh, I see. I hope that works. But what if it is not enough to solve all the problems?

Another problem.

http://www.librarything.nl/author/wolterkampmannmartin
Er/Daar zijn2keer besproken over/aangaande Martina Wolter-Kampmann's books.

A phrase cut into 5 parts is a real disaster. Please make it a single phrase with a plural and single form because of the number.

Sorry Tim, thanks to the latest "improver" I now can understand boekerij's despair ;-) But I think this thread is a lot more constructive than his way of communicating.

Okay. I fixed that one. Crazy, I agree. It now reads:

There are <SUB1> conversations about <SUB2>'s books.

Now, this is going to be a problem if either of the SUBs gets inflected, for example if the author's name should be in the genitive. What is the best approach here. LibraryThing is not about to start asking Dutch people to put every author into every grammatical case. You can, however, do something like:

There are <SUB1> conversations about this author.

That works fine.

In Dutch neither would be a problem. German might have a problem with the first version but that language was my nightmare at high school. So others should speak out.

There is another problem with

'There are SUB1 conversations about SUB2's books.':

The counter is wrong, or rather: the counter is right, but the link that should lead to the conversations only shows those where a title was used for the touchstone, not those where the touchstone was the author.

So 'There are SUB1 conversations about this author.' would be even more 'untrue'.

You can look at David Pogue (9 / 3) as an example - where you'll see this thread counted as no. 9, but no link to it.

What about

- There is one conversation about a book by this author.
- There are conversations about books by this author.

Translating the word "shared".

On the Shared Favorites page there should be possible to translate the word "shared" to singular and plural forms (1 shared, 2 or more shared).

>19 Anneli:

Done.

On Zeitgeist books page there is the word 'reviews':
25 most-reviewed books
The Da Vinci code (425 reviews)

In Finnish:
The Da Vinci code (425 arvostelut)

BUT that is wrong! It should be in partitive form:
The Da Vinci code (425 arvostelua)

If I change that, it also changes in the work's social page:
arvostelua (1)

There it should be in nominative form:
arvostelut (1)

Could this two be separated so they can be in different inflectional forms?

What a beautiful word, "arvostelut" :-)

What do you mean by "partitive form"? Why should the word bedifferent in both cases? I don't know a word of Finnish - or rather, I now know one - but I'm just curious.

>22 xtien:

xtien, there are fourteen or fifteen cases for nouns in Finnish. And we need them all! We don't use prepositions in Finnish, we use inflectional forms. I cannot explain why, but the partitive is used with numerals:

100 books = 100 kirjaa
BUT: 1 book = 1 kirja

It gets more complicated when the noun is an object:
I own these books = Omistan nämä kirjat.
I own 100 kirjaa = Omistan 100 kirjaa.
I own 1 book = Omistan 1 kirjan.

You can see the Finnish cases in this table and how they are used:
http://www.cs.tut.fi/~jkorpela/finnish-cases.html

Things are actually not that simple because in that table is just one word. The inflections are different for example with words that end with consonant. If you are interested there is a site that is more thorough:
http://www.cc.jyu.fi/~pamakine/kieli/suomi/sijat/indexen.html

Here in LibraryThing one big problem is the genitive form. In English you just put 's after the name and there is your genitive, but in Finnish the genitive is different depending on inflectional stem. My name is simple:
Anneli's = Annelin (just add 'n')
but there are names like Hilkka:
Hilkka's = Hilkan
or Aleksis Kivi:
Aleksis Kivi's = Aleksis Kiven.

>14 kantelier:

Die sneer erbij -- trappen naar een persoon, toch? -- was natuurlijk heel erg nodig?

Tim :

The only problem we had with the "There are n conversations about N.N.'s books" (combination) phrase was that the "N.N.'s books" part was not available for translation.

Present problems with LT.nl are manyfold. LT.nl has been ruined and turned into disaster indeed.

As far as I understand, the latter is the way you wanted it. Be assured that "correctors"--that, by the way, didn't intend to use LT.nl themselves, ever--have succeeded in turning it into a mess.

I am quite sorry that I do not think that this is funny. The more while I have been quite an avid user and supporter of nl.LT and LT.nl.

Please call me names. Others are doing so, too, anyway--AFAIK with your continued supporting them in this matter.

Your obliging doormat.

Anneli, how do you guys manage to learn Finnish at a young age? It's beautiful, but complex. I bet Finnish is a good language for poetry, right?

Your obliging doormat

Self knowledge is always a good thing ....

But, seriously, why are you doing this to yourself? If ten people vote for one translation and you change it into something else that's not even in the dictionary, you know you're not going to get a lot of sympathy.

And leave Tim out of it. Tim doesn't even speak Dutch.

Your remark on "not using LT.nl" doesn't make sense. Most of those who read English, use LT.com. Those who don't read English well, use LT.nl. You're not saying that those who don't speak English, should do the translation?

I vote for the creation of Libraryting.be. I will never read it, I will not contribute one word to the translation, I promise I will never criticize one bit of it.

btw, boekerij, on a completely different subject, I take it you have been on vacation for a few weeks? Haven't seen you around for a while. How was your vacation?

boekerij: I am sorry you see things in these terms. There is no conspiracy against you, or with "them." Nor are all the people you disagree with, myself included, actively trying to destroy the site. I wish you could see how silly these claims must look to others--that I'm actively trying to hurt the site by not letting you win every time.

As you should know, I have no opinions except that there be mechanisms in place which force participation and compromise, and which do not allow a small or singular minority to defy all others. I have asked you point blank what mechanism you thought would work better. You declined to respond. I would be happy to entertain and discuss them now.

I would add that progress on the translations has happened because other users started discussing the issues—your issues, in fairness—without constant harranging.

>27 xtien:

In Boekerij's defense, I take that to be rather snide. Let's try to reduce the temperature a bit.

>29 timspalding:
Tim, I apologize, I didn't mean it to be snide. It's vacation time in Holland and Belgium, everybody goes either south or north for three weeks, so I assumed that boekerij had been away. Not asking "how was your vacation" is rude. Even I have been on vacation for a couple of days, my kids came over and we went to Las Vegas, to Grand Canyon and then to Santa Cruz, for surfing. The latter was best :-)

Yes, you Europeans and your long vacations! How terrible.

>25 xtien:

xtien, if people manage to learn Dutch at early age, so why not Finnish ;-)

I agree, Finnish is the most beautiful of languages, like all mother tongues.

>32 Anneli:

Well, Dutch is much easier, I learned that without putting in noticable effort, before I went to kindergarten. French and German were much more difficult, took me years and years of hard studying before I was more or less fluent in those. So I don't want to think about Finnish......

There are problems with the word combined
for example on the helpers' log pages:

work combinations in English:
Anneli combined Forsytein taru by John Galsworthy, The Forsyte saga by John Galsworthy

in Finnish it is incorrectly:
Anneli yhdistetty Forsytein taru by John Galsworthy, The Forsyte saga by John Galsworthy

this should be:
Anneli yhdisti teokset: Forsytein taru / John Galsworthy, The Forsyte saga / John Galsworthy

Yhdistetty is passive particible and here is somebody doing something, so the verb should be in active personal form.

Maybe the simplest solution would be to use passive in all types of combinations:

combined: tag Evolution and tag evolution (username)
combined: author Robert Lee Brewer and Robert Brewer (username)
combined: work Forsytein taru / John Galsworthy and work The Forsyte saga / John Galsworthy (username)

or something like that.

Thanks. Instead of keeping an isolated "combined" the phrase is now:

combined (see work)

That should fix that. If combined is elsewhere wrong, let me know?

T

>35 timspalding:
Thanks, Tim. It works on author combination page:
http://fi.librarything.com/log_helpers.php?view=authors

but not on work combination page:

skittles yhdistettyLet's Talk About Being Wasteful (Let's Talk About Series) by Joy Wilt Berry, Being wasteful (Let's talk about) (Let's talk about) by Joy Wilt Berry (katso teos)

Sorry. I'm not quite getting what's not working can you explain closer.

Can you tell me what's wrong between:
http://www.librarything.com/combine.php?author=thomasdave
http://fi.librarything.com/combine.php?author=thomasdave

?

>37 timspalding:

I'm sorry, my explanation was inaccurate. I didn't mean those combination pages, but those on the helpers' log:

Zeitgeist -> Helpers -> Helpers log: Work combinations and Helpers log: Author combinations

Now I'm leaving on vacation: three weeks in Lapland. Without a computer, so I stop pestering you for a while. Or if I get withdrawal symptoms, I try to find a public library where I can use a public computer...

Wow. Lapland. Enjoy!

This is not a homograph but related issue anyway...

There should be singular and plural forms of the word copy on the combination page where works of an author are combined, for example this one:

http://www.librarything.com/combine.php?author=devillerfrdrique&sort=1

Fixed. (Well, give it a sec. I'm waiting to push a whole bunch of minor fixes together.)

This is a problem with the Hungarian version... (agglutinative languages for teh win :P )

On author pages, "books by" should be different here:

173 LibraryThing users own 267 books by Andrey Kurkov.

than here:

You have books by Andrey Kurkov

Currently they use the same field. One should be *all users'* books by and one should be *your* books by. Otherwise the author page looks completely mangled :(

Work-info page, e.g. this one:
Kuinka kaikki toimii (The way things work)

Owned by 470 users, 1 reviews

there should be both singular and plural forms of the words user and review

In Finnish language we need to have singular, partitive form with numerals (>1 timspalding:): 470 käyttäjää, 1 käyttäjä, so the word user=käyttäjä cannot be used in every context. The users of LibraryThing (all of them) = LibraryThingin käyttäjät, but two users of LibraryThing = kaksi LibraryThingin käyttäjää.

The menu bar of Zeitgeist:
overview | books | authors | tags | users | by language | helpers

should be in Finnish (the nominative forms):
yleiskatsaus | kirjat | tekijät | avainsanat | käyttäjät | kieliryhmittäin | avustajat
but the works authors (tekijät) and users (käyttäjät) are used in some other context, too, and so changing those in the menu bar cause trouble elsewhere.

Message to self: Tackle 42ff. (Sorry. I can't do either page right now; they're being worked on.)

> 42, 44

Color me confused, is the "you have X books by Jane Doe" sentence now completely removed from author pages? Because I can't see it, but maybe I'm looking in the wrong place... I'm running on a lack of sleep :O Anyway, it looks all right now.

But... I found another homograph on the author page (NO!! PLEASE DON'T HATE ME!) - it doesn't show if the author doesn't have too many books, which is why I didn't spot it for the first time.

Eg. here, "show all N books" only has "show all" as an editable field and I suppose it gets "books" from a "books" entry somewhere else. However, in Hungarian, "books" should be in accusative case here, while the default entry is in nominative. So right now it says "mutasd mind a N könyvek", while it should be "mutasd mind a N könyvet".

Unfortunately this cannot be fixed by paraphrasing "show all" in some slightly-contrived-but-still-understandable way, because in Hungarian, a number specifying a quantity is never followed by a plural noun. So "N books" is wrong regardless of the case (so we have two problems here, a case problem and a sg/pl problem).

This book/books thing was a constant problem in the Hungarian version, but I think it now displays all right everywhere else.

Maybe a solution could be to have a "show all books" kind of field instead of just "show all" + N + "books" - that way everyone could suffix/prefix/infix that noun to their hearts' content.

Sorry, accidentally posted the same thing twice...

On work page e.g. The diaries of Paul Klee, 1898-1918:

Recently added by

In Finnish we don't use such expressions. We need to have an expression where one person acitively add or persons add books or whatever. I translated recently added by like this:
Viimeisin tallentaja/viimeisimmät tallentajat (Most recent "adder"/"adders")

It would be better if there were singular and plural forms.

The word date (noun) cannot always be translated with one and the same word in Finnish (and there are probably some other languages as well):

date of publication (usually the year when the book is published) is in Finnish "julkaisuvuosi" or "julkaisuaika"

calendar date (e.g. 16 November 2003) is in Finnish "päivämäärä" or "päiväys"

If I translate "date" as "päivämäärä" it looks odd in book information. And if I translate it "julkaisuaika", it looks out of place in e.g. member review dates.

I was reading these messages, especially by Anneli, and I realize how difficult it can be to translate some sentences that have a context to languages like Finnish or other Uralic languages, Altaic languages, Caucasian languages or others that have so many suffixes. The so-called agglutinative languages.

I get the same problem as Anneli while trying to translate some words to Georgian. There are sometimes more than one word for different contexts or as has been discussed here in July 2007 the nouns have several cases and postpositions that make it hard to decide which one is best.

> 50

Paal :

If you are actively translating (or trying to), it might be a good idea if you can be reached somehow. Quite often, translators sent each other private remarks, asking and explaining about maybe different views, in order to gain mutual agreement.

What's more, if you want to change a text snippet translation that has received multiple cheers, it might be a good idea to double check and/or try and ask first, in order to prevent that your intended correction gets wrong.

> 16

Sorry, but, contrary to kantelier's "In Dutch neither would be a problem.", in Dutch, there definitely is a problem with the form :

There are conversations about 's books.

for Dutch language genitive form of e.g. kantelier is kanteliers, as in "kanteliers boeken", but genitive form of boekerij is boekerij's, as in "boekerij's boeken". Mark the former takes no apostrophe, but the latter does. For completeness, Dutch language genitive form of Jos is Jos', as in "Jos' boeken".

We have got a workaround indeed, by using the roundabout way.

A major problem with the latter though is that some (newly arrived) avid corrector that doesn't understand the workaround's purpuse, neither what way the text snippet is used, might "correct" it as he sees fit. Never mind his good intentions, the result might be less than satisfying indeed.

This is a really bad one:

to
from

are used on invite page as receiver and sender:
Tell people about LibraryThing

and on LT Local they are used as
starting time and ending time of events

Impossible to translate these with words that would fit in both places. Please, help!

Nobody probably reads this topic, but here is one really funny:
LibraryThing thinks you probably will like Nada : a novel (certainty: keskikoinen)

(certainty: keskikoinen) is in English certainty: average size.

I think that certainty is expressed as very low, low, medium, high or very high. But it seems that the same words has been used to express something else, too. Oh, well I change the translation, but I wish I knew where else (in what context) these words has been used previously.

> Nobody probably reads this topic
I do :)

I noticed similar issue in Polish version:

"certainty: bardzo wysokie"

"Certainty" in Polish is "pewność". And the problem here is that it should be:

"pewność: bardzo wysoka" - pewność is a feminine noun and requires feminine form of adjective while "bardzo wysokie" is a neuter adjective.

For now I translated "certainty" as "prawdopodobieństwo" (probability) so that the grammatical gender is correct...

The English word title has many meanings and cannot be translated with just one expression.

Translation of "title" cannot be the same as book title or work title in this context (widget title):
http://www.librarything.com/ltsw/

The same "title" seems to be in use in this page also (I am not sure):
http://www.librarything.com/er_pending.php?status=save&program=giveaway

The word subject has more than one meaning. In LibraryThing at least two meanings has to be separated: subject heading and talk topic subject. I just noticed that if I start a new topic, the subject line in Finnish is subject heading. I cannot seem find where to change that. I must go to the All translated page and that is terrible.

Gah!

"unread" on, for example, the home page overview of Talk*, is plural, but the same string is used in the actual topic for "unread" on each message† is singular. Almost every language must have a problem with this.

* as in:

New features : Author Division: Philosophy 18 unread / 40

† as in:

unread oct 9, 2010, 5:22am (up)

This is not a homograph, but a word with slightly different meanings, which necessitates different translations for many languages but with only one entry: Events.

CK events (historical events) are not the same as events that your local bookshop arranges. But that English single string is used in both places. Problem for Swedish and most likely many more.

*groan*

You are using the same "Add" string both for "Add" buttons (for example, when you assign a contact to a list) and for filtering "adds" in the helper logs. In the first case it's a verb, in the second case it's a noun. This is impossible for the rest of the world to translate.

On book page the word possible must have singular and plural forms in translations:

Other authors
» Add/edit other authors (2 possible)

>61 Anneli:

Fixed.

From now I won't be reading this thread. Post bug reports separately. Thank you.

Duly noted!