Feminization hell
Moderator: Forum Moderators
Feminization hell
Until recently, there were only a few female units in BfW, so I tried to ignore this issue. However now, when female units of all kinds appear everywhere, I must scream
Untranslatable madness!
In Slavic languages, at least, gender is an ubiquitous grammatical category. Not only masculine, feminine, and neuter are (usually) different for nouns and pronouns -- so male fighter and female fighter must have different names, but adjecitves have gender too, so strong, quick, poisoned, etc. must have different translations for males and females (and neuters), as well as many verbs and numerals, although these two are less problem in BfW.
If this is going to be resolved somehow and players and translators are not going to be left in the cold with unfixable, agrammatical, odd phrases, please also take into account that some languages (like mine) are crazy enough to assign arbitrary genders to arbitrary nouns -- so while tentacle and wraith are neuters (in Czech), spider and scorpion are masculines, and cuttlefish and galleon are feminines. Other languages can assign other arbitrary genders. So, ideally, translators would define the gender of a unit [for the translation purpose].
Frankly speaking, I don't see a point in continuing Czech translation, if this issue will not be resolved.
Untranslatable madness!
In Slavic languages, at least, gender is an ubiquitous grammatical category. Not only masculine, feminine, and neuter are (usually) different for nouns and pronouns -- so male fighter and female fighter must have different names, but adjecitves have gender too, so strong, quick, poisoned, etc. must have different translations for males and females (and neuters), as well as many verbs and numerals, although these two are less problem in BfW.
If this is going to be resolved somehow and players and translators are not going to be left in the cold with unfixable, agrammatical, odd phrases, please also take into account that some languages (like mine) are crazy enough to assign arbitrary genders to arbitrary nouns -- so while tentacle and wraith are neuters (in Czech), spider and scorpion are masculines, and cuttlefish and galleon are feminines. Other languages can assign other arbitrary genders. So, ideally, translators would define the gender of a unit [for the translation purpose].
Frankly speaking, I don't see a point in continuing Czech translation, if this issue will not be resolved.
-std::string font_name = "Vera.ttf";
+std::string font_name = "Bepa-Roman.ttf";
+std::string font_name = "Bepa-Roman.ttf";
Well, I'll start with a short sketch of what I think is needed, and then translators from other languages will have to fill in what I've missed.
(Note: this is not based on what swedish needs, but what a basic adjective-congruent-with-noun-gender language might need)
*The translator needs to be able to specify an arbitrary (but small) number of grammatical genders [f.ex male female neutral...] and associate each affected translation with forms for each gr. gender.
* For all non-gender divided units; i.e., it (cuttlefish, f ex), the translator needs to specify which gr. gender this unit should use.
*For all gender divided units; the translator needs to be able to specify which gr. gender to be used for each natural gender. (while this ususally is natural gender = female, then gr. gender also = female, this is not always the case). However, many languages will allow adjectives congruent with natural gender, at least in a list context, such as the right menu, so perhaps in this case we may assume natural gender = grammatical gender?

*The translator needs to be able to specify an arbitrary (but small) number of grammatical genders [f.ex male female neutral...] and associate each affected translation with forms for each gr. gender.
* For all non-gender divided units; i.e., it (cuttlefish, f ex), the translator needs to specify which gr. gender this unit should use.
*For all gender divided units; the translator needs to be able to specify which gr. gender to be used for each natural gender. (while this ususally is natural gender = female, then gr. gender also = female, this is not always the case). However, many languages will allow adjectives congruent with natural gender, at least in a list context, such as the right menu, so perhaps in this case we may assume natural gender = grammatical gender?
Hello,
The idea is interesting for french too.
* In german, the sun is a female noun and the moon a male one.
* In french, the sun is a male noun and the moon a female one.
Just my two cents,
Le Gnome
The idea is interesting for french too.
I have an example :sanna wrote: *For all gender divided units; the translator needs to be able to specify which gr. gender to be used for each natural gender. (while this ususally is natural gender = female, then gr. gender also = female, this is not always the case). However, many languages will allow adjectives congruent with natural gender, at least in a list context, such as the right menu, so perhaps in this case we may assume natural gender = grammatical gender?
* In german, the sun is a female noun and the moon a male one.
* In french, the sun is a male noun and the moon a female one.
Just my two cents,
Le Gnome
Hm, I'm back, but sanna has already written almost exactly what I would. I just don't know whether gender differentiated units, i.e. humans (I count elves, etc. in too here), need to define grammatical gender too, I was hoping all languages either don't differentiate between genders or have grammatical and factual gender the same for human beings. Maybe it's just a lack of imagination...Dave wrote:Can you give me a list of all the changes that will need to be made to support this? We'll scope out the difficulty.
I'll add what I feel it needs to depend on gender most badly now (this is strongly affected by me being Czech ;-)
- the names of gender differentiated units themselves -- i.e., allow different translation of male fighter and female fighter, male mage and female mage, etc.
- adjectives that describe unit properties -- traits, slowed, poisoned, maybe abilities, even though I was able to write these in a gender-neutral form IIRC
And I'll also add a list of a few other things that may or may not interfere, so one should be aware of them (I do not request to fix them immediately ;-)
- Plural forms -- I hope this one will not interfere. There's a lengthy chapter about this topic in gettext manual, English IMHO could benefit too, although some messages are already special-cased for English, like `$friends Friendly units sighted'/`Friendly unit sighted' (but this may not be enough for some languages, see `Plural forms' in gettext manual). Using the existing gettext support would help a lot.
- Inflexion -- there's definitely some interference, but this problem is deeper and IMHO too much pain for too little gain, i.e., insolvable. Some cases are inherently hardly solvable, like `$login's game' (an algorithm transforming a noun to a corresponding adjective for a one particular language may have a few pages and still get some exceptions wrong), this is translated as `$login plays' in Czech for example (not very nice, but makes some sense). Some are like `Do you really want to dismiss $noun' or `they must have captured $R2.user_description' that I have to write in a form, where nominative of any gender can be used in place of $something, or just leave out the variable and write it generally. As I said, I don't really expect any `fix' here.
- Capitalization -- this interferes, but in a simple `everything 2x' manner in the worst case I hope. The problem is that English allows you to apply capitalization quite liberally, so you have `Elvish Hero' and use it everywhere, but I can't. I should write `Elfàhrdina' in standalone labels and at the start of a sentence, but `elfàhrdina' in the middle of a sentence. This interferes badly with the previous point, but I just ignore it for now.
Last edited by yeti on September 2nd, 2004, 7:55 pm, edited 1 time in total.
-std::string font_name = "Vera.ttf";
+std::string font_name = "Bepa-Roman.ttf";
+std::string font_name = "Bepa-Roman.ttf";
This is something that has irritated me for quite a while, translating unit names with lower case would yield annoying results when they're actually used sentence initially.yeti wrote:- Capitalization -- this interferes, but in a simple `everything 2x' manner in the worst case I hope. The problem is that English allows you to apply capitalization quite liberally, so you have `Elvish Hero' and use it everywhere, but I can't. I should write `Elfàhrdina' in standalone labels and at the start of a sentence, but `elfàhrdina' in the middle of a sentence. This interferes badly with the previous point, but I just ignore it for now.
Would it be possible to allow translators to use lower case, and have it capitalized in certain circumstance; like sentence initally and as stand alone. Languages like english that can/should use upper case much more frequently could then use those capital letters in their translation; i.e., initial lower case letters can be capitalized in certain environments, but upper case letters never modified.
This is of course only feasible if we can come to a consensus of a limited number of positions in which to capitalize them.
However, while the unwarranted upper case letters 'look ugly' in languages that don't use them, they do not severly interfere with game enjoyment, so fixing this should not be a top priority.
So, what do we do
So, is there going to be any resolution about this issue. I am currently translating into Bulgarian and when i reach the point with the units and unit descriptions things are going to get messy for me too. As in all Slavics i have the same problems as decribed above. The few strategy games that are developed in Bulgaria or have Bulgarian translations do not include the concept of Female units. There are some workarounds but trust me, most of the things are not going to match well and will look ridiculous at some points.