1079 réponses à ce sujet

[Wissenschaft]

Xewaka wrote...

Wissenschaft wrote...

Or maybe Graunt got so immersed in Mass Effect they he felt his Shepard was his own character.

Inmersion and personalization have nothing to do with each other. I can be inmersed in a good novel, yet I won't find myself appropiating Vimes' persona.

That doesn't mean Graunt felt the same way. Quite possible he felt the opposite of you.


As for paraphrasing, I'd rather them use better translations than abandon paraphrasing. It seems like it fits very well with the type of story bioware is telling in DA2.

Modifié par Wissenschaft, 25 décembre 2010 - 01:38 .

[Xewaka]

Wissenschaft wrote...
As for paraphrasing, I'd rather them use better translations than abandon paraphrasing.

The point I am trying to get across is that paraphrases, by their very nature, do not allow good translations.

[upsettingshorts]

I'm willing to believe that if they must be translated in isolation from their resulting dialogue and action. Which seems wildly counter-intuitive to me.



I just don't see any reason at all why they must be translated in that fashion.

[Xewaka]

Upsettingshorts wrote...
I'm willing to believe that if they must be translated in isolation from their resulting dialogue and action. Which seems wildly counter-intuitive to me.
I just don't see any reason at all why they must be translated in that fashion.

Me neither. But that's how files are handled.

[Maria Caliban]

HolyJellyfish wrote...

History itself is built on unreliable narration.

The term 'unreliable narrator' is receiving a great deal of abuse. It's a literary technique. While it has something to do with human nature, it doesn't appear in the real world because reality is not narrated.

If I, Maria, believe something completely and tell it to you, I am not an unreliable narrator because I'm a real person. If I create and narrate documentary about how government fluoridation in the water has directly lead to the increase of cancers, and I totally believe this, it doesn't matter whether what I say is the truth or not. Even if I don't believe everything that I put in the film but am exaggerating or misrepresenting the facts so to make my position more persuasive.

However, imagine that don't believe that fluoridation is causing cancer, but pretend to believe it so I can show you a narrator who is completely paranoid about something. I am using this as a broader critique of how paranoia impacts American politics and culture. I have "Maria" ambush a congressman from Alaska, and demand to know what he thinks about Mulnoma County introducing fluoride to the drinking water? AND does he know that fluoride consumption was linked to cancer when consumed during formative years? AND does he know that fluoride is a waste product of the aluminum industry and over 70% of the funds for pro-fluoridation advertizing in Mulnoma County came from an aluminum manufacture?

The guy starts sweating because I just got in his face with a camera, am talking about something he's never heard of, and it apparently causes cancer and there are special interest groups involved. Maybe he runs off or maybe he blusters on about stopping the aluminum industry, protecting our children, etc.

The idea here is that "Maria" obviously doesn't have all her marbles. We learn that "Maria's" mother, grandmother, aunt, and two of her cousins have died of cancer, and we see her relentlessly purifying her water in an attempt to get all the fluoride out and save herself. This fear has obviously warped her perception and she's focused on a 'cause' she can control and eliminate.

The unreliable narrator isn't simply lying or exaggerating (there are many literary traditions that do this and they pre-date the UN by quite a bit*) it's about narrative tension. The audience learns to doubt the content of the story, and they're often forced to understand the narrator to separate the real bits from the false bits. They might ultimately feel that they see beyond the biases (The Book of New Sun) or they might ultimately feel that they can never know the true story (Roshimon).

*The song linked in my sig, Tribute, is about two brothers who meet a demon and manage to save themselves by singing the greatest song in the world. 'How I met a devil and got away' is a popular plot in Christian folklore (also see, the Devil Went Down to Georgia). It's not fantasy, the speaker and audience presumably believe in demons, devil, and angels, but it a tall tale.

Modifié par Maria Caliban, 25 décembre 2010 - 02:33 .

[In Exile]

Xewaka wrote...

Yet this failure doesn't happen with full text. No more than regular print typos, anyway.

I can't comment on that because I have never played a translated RPG. I suppose JRPGs might count, but I had an ex-gf who was really into anime and I picked up enough Japanese to know when the subtitle and line don't actually match, which was pretty often. The honorific and verbal tics is lost completely. You maybe get 60% of the intended content of the speech in a translation.

I've done freelance translation. I know how hard it is to rewrite short lines of text to have them make sense in a different language while keeping the intended meaning and the character count limit. Paraphrases simply don't give you neither enough character count nor enough meaning to work with them.

I can't comment on whether or not the character limit is restrictive. I have no experience with this, so I can't speak to the accuracy of this statement. What I can say is that separating the paraphrase from the actual text is stupid, because the dialogue & paraphrase are written in conjuction. That isn't a problem with the paraphrase - that's a problem of idiotic localization.

Like I said: a paraphrase ought to never be translated. How it should be done is that the full text needs to be translated and then a paraphrase written for it (likely how it works for the English verison too).

So if localizations are near flawless with a method yet completely unreliable with another, I'd say the blame is in the method, not the localization.The problem comes clearly from the paraphrase.

No, the problem comes from the method. You are assuming that the method is universally useful.

That's wrong. If I move faster when I pedal on a bike, and then I swap my bike for a car, complaining the car has a problem because it lacks pedals is not a good criticism at all.

Not together.

Which goes right to the way localization being handled is idiotic.

Modifié par In Exile, 25 décembre 2010 - 02:32 .

[Ziggeh]

Addai67 wrote...

I'm interested in seeing how the framed narrative works out, but it does seem to add an extra layer of distance between the player and Hawke. If you were really playing Hawke, you would know what actually happened and wouldn't need Varric to tell the story. Maybe the trade-off will be worth it, though, especially for those who don't mind the VO and dialogue wheel.

I agree it's another layer, but I'm not sure it should be seen as a barrier. After all, you're already playing through a story that never really happened, it's only a small step to also contain playing through a story that may not have happened.

I'm fairly bias as I'm fascinated by fictions ability to construct and to question, but I can see how it might present RP problems, as in, if I attribute thoughts I'm having to Hawke, are they thoughts Hawke had or thoughts Varric said he had, or external to the whole process? I feel it's a questioning worth asking, even if you find an answer you might not like, and I'm actually a little sad I don't RP in the manner that many here do. My process will barely be affected at all, while it opens others up to all sorts of interesting reflection upon their approach.

[upsettingshorts]

I'd think about it like this:

You did play Hawke. In the past. Now Varric is describing how it went down to someone else, Cassandra. The fact that you're playing the game in the present tense and Varric describes it after isn't strictly in conflict with the game's narrative structure. That's distance, sure, but it isn't an unworkable one for an active imagination.

Modifié par Upsettingshorts, 25 décembre 2010 - 02:37 .

[Xewaka]

In Exile wrote...

Xewaka wrote...
I've done freelance translation. I know how hard it is to rewrite short lines of text to have them make sense in a different language while keeping the intended meaning and the character count limit. Paraphrases simply don't give you neither enough character count nor enough meaning to work with them.

I can't comment on whether or not the character limit is restrictive. I have no experience with this, so I can't speak to the accuracy of this statement. What I can say is that separating the paraphrase from the actual text is stupid, because the dialogue & paraphrase are written in conjuction. That isn't a problem with the paraphrase - that's a problem of idiotic localization.

Like I said: a paraphrase ought to never be translated. How it should be done is that the full text needs to be translated and then a paraphrase written for it (likely how it works for the English verison too).

Not together.

Which goes right to the way localization being handled is idiotic.

Keep in mind that this is not only translation. This is localization. You have to delve through the game data files to locate the text that will be displayed, then deformat it, translate it, reformat it, and reattach it to the data files. The localization of the paraphrases and the actual dub line aren't even close in the files. To attempt to rewrite the paraphrases would be even worse, as now you're basically recoding all the dialogs. And that isn't materially feasible.

Modifié par Xewaka, 25 décembre 2010 - 02:44 .

[upsettingshorts]

Xewaka wrote...

Keep in mind that this is not only translation. This is localization. You have to delve through the game data files to locate the text that will be displayed, then deformat it, translate it, reformat it, and reattach it to the data files. The localization of the paraphrases and the actual dub line aren't even close in the files. To attempt to rewrite the paraphrases would be even worse, as now you're basically recoding all the dialogs. And that isn't materially feasible.

Even if its not feasible, it's still not a problem with the paraphrase concept.  Hypothetically, if the paraphrase was next to its resulting full line or action - translation wouldn't be much more of an issue than it already is. 

The fact that it isn't can't be considered relevant when discussing the merits of the paraphrase within the game for that reason.  It's a localization problem.

Modifié par Upsettingshorts, 25 décembre 2010 - 02:49 .

[In Exile]

Upsettingshorts wrote...

I'd think about it like this:

You did play Hawke. In the past. Now Varric is describing how it went down to someone else, Cassandra. The fact that you're playing the game in the present tense and Varric describes it after isn't strictly in conflict with the game's narrative structure. That's distance, sure, but it isn't an unworkable one for an active imagination.

This is one of those subjective ''does it add a layer of separation or not'' issues like PC VO.

For those that feel that the hero is theirs, this is one of those things that actively works against that.

[Xewaka]

Upsettingshorts wrote...
Even if its not feasible, it's still not a problem with the paraphrase concept.  Hypothetically, if the paraphrase was next to its resulting full line or action - translation wouldn't be much more of an issue than it already is. 
The fact that it isn't can't be considered relevant when discussing the merits of the paraphrase within the game for that reason.  It's a localization problem.

Yet it is a problem that derives from insufficient information for the localization, and a problem that is easily solved with a proper full text choice, which does have the information. Thus the problem originates in the paraphrase, not the localization.

Modifié par Xewaka, 25 décembre 2010 - 02:54 .

[upsettingshorts]

In Exile wrote...

This is one of those subjective ''does it add a layer of separation or not'' issues like PC VO.

For those that feel that the hero is theirs, this is one of those things that actively works against that.

Yeah, I know. I was trying to offer a potential "workaround" of sorts by saying that in DA2 the hero was theirs at some imagined time in the past, and is Varric's now. It isn't all that far removed from figuring that your characters brush their teeth off camera, but I won't equate the two - it's not my place to draw the line as to what constitutes too much suspension of disbelief for everyone else.

Just had an idea of an approach to take for those concerned with it. It's how I'd do it. Doesn't mean they gotta.

Xewaka wrote...

Thus the problem originates in the paraphrase, not the localization.

No it doesn't. 

If the paraphrase was right next to its resulting full line or action there would be no problem.  The fact that it isn't is a problem of format, not content.   I wouldn't want to translate a paraphrase contained in one book that led to a line contained in a different book that was stored in a different room either, but in that case I'd probably want to move both books next to each other, as opposed to throwing one of them away claiming it was the problem.

Modifié par Upsettingshorts, 25 décembre 2010 - 03:01 .

[In Exile]

Xewaka wrote...
Keep in mind that this is not only translation. This is localization. You have to delve through the game data files to locate the text that will be displayed, then deformat it, translate it, reformat it, and reattach it to the data files. The localization of the paraphrases and the actual dub line aren't even close in the files. To attempt to rewrite the paraphrases would be even worse, as now you're basically recoding all the dialogs. And that isn't materially feasible.

That seems silly. A localized version is released at the same time as the English version, by the same company. I don't understand how that means they ahve to actually break into the game data files and can't just have direct access to the script.

[In Exile]

Xewaka wrote...
Yet it is a problem that derives from insufficient information for the localization, and a problem that is easily solved with a proper full text choice, which does have the information. Thus the problem originates in the paraphrase, not the localization.

No. The problem derives from some half-baked attempt at translating each line in isolation that might have worked when the script was done in a different way.

A technological innovation changes the requirement for implementation. That a problem comes up with a the implementation does not mean the old method was better; it means the implementation is now bad.

That's not to try to add objective value to a subjective thing (i.e. the paraphrase). But trying to pin the problem on the paraphrase is to say that there is some kind of objective value in the localization process being the way it is and that the dialogue should accomodate the localization scheme instead of the other way around, which is just silly.

[Xewaka]

In Exile wrote...
That seems silly. A localized version is released at the same time as the English version, by the same company. I don't understand how that means they ahve to actually break into the game data files and can't just have direct access to the script.

Translators are handed the raw text. If I'm not mistaken, there'll be an "interface" text block and a "subtitle" text block. The paraphrase is in interface. The actual line is in subtitle.
By the way, the example I made about problems with translating "back" in isolation - that there are two different words for it depending on if it refers the body part or returning? It actually happened in Warhammer Online for three weeks (I reported the translation bug, and I'm sure I wasn't the only one). Now try and imagine how much worse can it get when half the data is like that.

[upsettingshorts]

Until you present a scenario that says: "Even when the paraphrase and its resulting full line or action are literally right next to each other for the translator to see - the localization is too difficult and the resulting translation too inaccurate and unreliable" I'm going to have a real hard time - no, impossible time - accepting that the paraphrase concept itself is the problem here.

From where I sit, the problem you're describing tells me that the paraphrase concept calls for a different localization procedure. Not that the concept should be abandoned because the current procedure presents significant problems.

Modifié par Upsettingshorts, 25 décembre 2010 - 03:07 .

[tmp7704]

In Exile wrote...

But you haven't given any instances. At best you've given psychologically impossible scenarios like being able to attend to music without attending the coherent organizing framework for it.

Whether what i've provided is actually psychologically impossible is up for debate -- as it is, our argument is very much on preschool level with me saying some things are possible and you countering that no, they aren't. I base my statements on  how i perceive my personal experiences, on what exactly you're basing your opinion that what i'm experiencing is impossible, i don't know since you aren't really elaborating on it.
 

The framework is what makes it music and not noise to begin with.

The way it's put and with the earlier part of discussion in mind, i interpret your stance to be akin to saying one can't just like colour red e.g., they can only like things painted red. And they can only actually enjoy the thing itself, it being painted red is only modifier to level of their enjoyment.
 

That your opinion is entrenched isn't proof that it's true.

Granted of course, and it works both ways.

Modifié par tmp7704, 25 décembre 2010 - 03:18 .

[AlanC9]

Xewaka wrote...
I've done freelance translation. I know how hard it is to rewrite short lines of text to have them make sense in a different language while keeping the intended meaning and the character count limit. Paraphrases simply don't give you neither enough character count nor enough meaning to work with them. Full sentences do. So if localizations are near flawless with a method yet completely unreliable with another, I'd say the blame is in the method, not the localization.The problem comes clearly from the paraphrase.

What if they treated the paraphrases like a new problem?  Just ignore the English paraphrase and whoever's doing localization is responsible for writing a new paraphrase based on his understanding of the subsequent line

[In Exile]

tmp7704 wrote...
Whether what i've provided is actually psychologically impossible is up for debate -- as it is, our argument is very much on preschool level with me saying some things are possible and you countering that no, they aren't. I base my statements on  how i perceive my personal experiences, on what exactly you're basing your opinion that what i'm experiencing is impossible, i don't know since you aren't really elaborating on it.

When I speak about psychologically impossible, I put on my neuroscientist had and refer to empirical research. Like the writing finding. It is not possible for a literate person not to read something but in front of them if they actually look at it, in the same way it is impossible for them not to ''see'' something when it falls within their attentional spot
 
In this case, you've already argued that people can be entirely wrong about why they believe certain things, so if you grant me this, you've already granted me the only avenue I need to appeal to psychological findings.

The way it's put and with the earlier part of discussion in mind, i interpret your stance to be akin to saying one can't just like colour red e.g., they can only like things painted red. And they can only actually enjoy the thing itself, it being painted red is only modifier to level of their enjoyment.

No, because the colour red is separable from red paitings. This is like your icecream flavour analogy for companion customization - it pressuposes your POV.

[Maria Caliban]

In Exile wrote...

When I speak about psychologically impossible, I put on my neuroscientist had and refer to empirical research. Like the writing finding. It is not possible for a literate person not to read something but in front of them if they actually look at it, in the same way it is impossible for them not to ''see'' something when it falls within their attentional spot

I'm going to counter with the rather simple experiment where you have the word BLUE in red colors on a white flashcard. You flip the flashcard and have someone immediately state what color the word is.

Most people will say 'Blue.' Some will not. People who originally say 'blue' can learn to ignore the word and state the color instead.

The human brain excels at ignoring what is 'unimportant' and focusing on what's 'important,' but it first has to learn what is important and unimportant. It's much like the experiment where a gorilla appears in the middle of a group of basketball players - some people never see it and those who do see it don't understand how anyone could miss it.

It is certainly not impossible to ignore text in a language you understand. It's difficult for some and impossible for some.

What about the text that scrolls at the bottom of a newscast while the anchor is talking? I usually see it at first and then ignore it because the 'important' stuff is what the newsperson is saying. Also, it's a human face and voice; that's far more engaging to me than text. After the first minute of watching, if you asked me what the text was, I'd have no way of telling you.

[In Exile]

ETA:

Just to correct some misconceptions, since we did talk science for a page I will not delete this post, but I will not post anything further to derail this thread as David said.

Maria Caliban wrote..

I'm going to counter with the rather simple experiment where you have the word BLUE in red colors on a white flashcard. You flip the flashcard and have someone immediately state what color the word is.

Most people will say 'Blue.' Some will not. People who originally say 'blue' can learn to ignore the word and state the color instead.

Are you talking about practice effects for the Stroop effect?

You can demarcate that into two groups. The ones that do say 'red' say it with a statistically significant delay that shows content monitoring (i.e. they read blue, then consciously filter it, and we can measure this via EEG).

The human brain excels at ignoring what is 'unimportant' and focusing on what's 'important,' but it first has to learn what is important and unimportant. It's much like the experiment where a gorilla appears in the middle of a group of basketball players - some people never see it and those who do see it don't understand how anyone could miss it.

That's attentional scaling, but attentional scaling is very different from automated processing.

It is certainly not impossible to ignore text in a language you understand. It's difficult for some and impossible for some.

Not ignore. Rather,  ''not read''. This is what we are talking about. There is a distinction. In the Stroop effect, we measure the interference effect of reading on the automaticity of the response. This isn't the same as demonstrating they are not reading the word.

There is excellent evidence available to show that people do automatically read the word in the Stroop effect. The significant difference in response is not relate to a capacity to avoid reading but rather to a more powerful cognitive supression effect and more powerful response monitor.

What about the text that scrolls at the bottom of a newscast while the anchor is talking? I usually see it at first and then ignore it because the 'important' stuff is what the newsperson is saying. Also, it's a human face and voice; that's far more engaging to me than text. After the first minute of watching, if you asked me what the text was, I'd have no way of telling you.

That's just attentional scaling, again. Showing that attention is a spotlight does not disprove automaticity in reading.

Modifié par In Exile, 25 décembre 2010 - 09:14 .

[Wissenschaft]

Xewaka wrote...

Upsettingshorts wrote...
Even if its not feasible, it's still not a problem with the paraphrase concept.  Hypothetically, if the paraphrase was next to its resulting full line or action - translation wouldn't be much more of an issue than it already is. 
The fact that it isn't can't be considered relevant when discussing the merits of the paraphrase within the game for that reason.  It's a localization problem.

Yet it is a problem that derives from insufficient information for the localization, and a problem that is easily solved with a proper full text choice, which does have the information. Thus the problem originates in the paraphrase, not the localization.

I'd rather the game only be avalible in english and keep the paraphrasing.

Well see how well the localization is handled when DA2 comes out but localization problems are not a reason to abandon the use of paraphrasing.

Modifié par Wissenschaft, 25 décembre 2010 - 03:50 .

[Maria Caliban]

In Exile wrote...

That's just attentional scaling, again. Showing that attention is a spotlight does not disprove automaticity in reading.

I think I see where the conflict is. I view a certain level attention scaling as effective ignorance where you need evidence of absolute unawareness?

[tmp7704]

Maria Caliban wrote...

I'm going to counter with the rather simple experiment where you have the word BLUE in red colors on a white flashcard. You flip the flashcard and have someone immediately state what color the word is.

Most people will say 'Blue.' Some will not. People who originally say 'blue' can learn to ignore the word and state the color instead.

The human brain excels at ignoring what is 'unimportant' and focusing on what's 'important,' but it first has to learn what is important and unimportant.

Interesting; i was scrolling this thread up from the end of it to catch up, so i just skimmed across the lines. Without knowing the context, the BLUE thing registered as 'something red' in my head and "blue" very much failed to register at all. I don't think there was any delay in the process.

I suppose it may put me in that group of "some who will not". Maybe that's why i have this trouble with finding common ground with In Exile in this discussion -- he literally perceives things in different way and focuses on different aspects.