Language System v2 feedback

[Ithalan]

A couple of observations:

• The new inline language assignment using brackets (for example [chondathan] Text to say in Chondathan) is great, but the way it is rendered in the chat log leaves a bit to be desired when you just want to leave individual Nouns understandable to everyone by having them treated as Common. Saying [chondathan] Hello there, my name is [common] Ithalan. [chondathan] How do you do? renders as [chondathan] "Hello there, my name is" [common] "Ithalan." [chondathan] "How do you do?". Which frankly looks a bit silly.
  
  Having a set of characters, for example <>, that mimics the old functionality of just displaying a particular string of text literally regardless of whether the rest is garbled or not, would be great.
  
  
• The patch notes seemed to indicate that [] would also work as a shorthand for typing [common], but this does not appear to be the case. Having the short-hand available (and shortened versions of other language tags in general) would be great for making them easier to use during play.

 

[RealZeratul]

Especially for higher levels of proficiency, maybe garbling of words could be changed to anagrams, exchange of vowels, adding or stripping trailing consonants (e.g., plural "s"), etc.

One of the often-heard points of criticism is how cumbersome it is to ask for word repetition, and that a native speaker should be able to understand it, anyway. With the last part I agree only to a certain degree (beginner-level speakers should probably not talk about the intrinsics of thaumaturgical philosophy), but this instead of full garbling should alleviate this perceived issue.

(also please make halfling DM gated, it's supposed to be secretive )

 

[anthee]

OK, here's an idea that would retain the basic premises behind the system but make it less intrusive and less immersion-breaking:

Since the language actually used is always English, you could script in mistakes which are typical for English learners. This would take some linguistic expertise but the DMs need but ask if they want the input of someone who has actually studied linguistics. (I'm looking at you, Kitty, but I'm sure you're not the only one!) A relatively easy way to do this would be to introduce phonetic mistakes based on English phonetics. I know next to nothing about scripting but if garbling individual words is possible, I don't see why changing/adding individual letters would not be possible in the same way.

An example:

'S' denotes, always and in every case in English, a sibilant sound. Other sibilants distinct from the possible values of 'S' – sounding similar but still different, in other words – are denoted by e.g. 'Z' and 'SH'. One possible mistake would involve shifting letters within the same phonetic group, like sibilants in this case, so that somebody who intended to say "You shall see" ended up saying for example "You sall zee". The mistakes there are obvious but the sentence remains intelligible, as sentences in real life often still do even when mistakes are made.

If enough possibilities like the above one were scripted in, the result would likely feel much more realistic and much less annoying. I know this is not going to satisfy those of you who feel that the chat box is sacred and holy, but if the basics of the current system are going to be kept, this would be one way to improve it.

Grammar mistakes would be a lot trickier, and on a grand scale, they'd be impossible to implement because it would take a level of code similar to the analytics of machine translators like Google Translate. But some things would be possible even then: for example, removing articles (a/an/the) should be pretty easy and it would simulate a very common mistake by English learners speaking a non-Germanic, non-Italic language as their mother tongue.

That's that. Don't shoot me please! I think the system can be made to work with some thought and effort.

 

[Ithalan]

Especially for higher levels of proficiency, maybe garbling of words could be changed to anagrams, exchange of vowels, adding or stripping trailing consonants (e.g., plural s), etc.

One of the often-heard points of criticism is how cumbersome it is to ask for word repetition, and that a native speaker should be able to understand it, anyway. With the last part I agree only to a certain degree (beginner-level speakers should probably not talk about the intrinsics of thaumaturgical philosophy), but this instead of full garbling should alleviate this perceived issue.

Another alternative to garbling might also be to just have some of the text get treated as if it was spoken in common, to reflect the speaker falling back onto the trade tongue for unfamiliar words rather than speaking incomprehensible gibberish. I feel like that would work better for keeping the conversation going while still clearly communicating that the speaker is not fluent.

(as a bonus, this would also make it possible for fluent speakers to fake not being fluent)

 

[RealZeratul]

Another alternative to garbling might also be to just have some of the text get treated as if it was spoken in common, to reflect the speaker falling back onto the trade tongue for unfamiliar words rather than speaking incomprehensible gibberish. I feel like that would work better for keeping the conversation going while still clearly communicating that the speaker is not fluent.

(as a bonus, this would also make it possible for fluent speaker to fake not being fluent)

I'm not strictly opposed to that idea, but I see two drawbacks:

• Aas languages are only conveyed via tags, it could read a bit cumbersome, as in "[Chondathan] Hello, [Common] this [Chondathan] is [Common] a [Chondathan] simple [Common] sentence",
  compared to, e.g., "[Chondathan] Hello, iths is o simple snectene".
  Granted, snectene is fairly hard to decipher, but one could count every single letter switch as a single obfuscation, so high skill means very few switches, and low skill makes it very hard or even impossible to decipher.
• It would make it very dangerous to hold secret conversations in rare languages, although with the learning mechanic, that's probably pretty dangerous, anyway.

 

[KiTTY]

As of 17th June:

With the old system before the big patch, words and phrases in Common blended seamlessly into conversation, and you could also have part of a word in Common, some[what] like this, whereas now, there's an obtrusive [Common] tag, and every changed language forces a new sentence. I suggest removing this tag when it's Common and bracketing the text in place, maybe with something other than [square brackets], since this feature is most often used with names. I'm less sure about how to deal with the autoformatting forcing a new sentence with auto-capitalization.