Latinized representations of sounds
I want to suggest those be eliminated, and that lessons rely entirely on matching audio to script.
There are many sounds in Hindi and Urdu that just don't lend themselves to representation with Latin characters (like the variety of "t" sounds) and even where it is possible, it isn't accomplished well. The use of a bar over a Latin character isn't sufficient to mark the many unique vowel sounds. On other occasions, I think the choice of Latin character is just wrong. (ca in my mind doesn't represent the sound "cha".)
The choice of Latin characters is actually not arbitrary: it is an ISO standard (ISO 15919), close to the IAST scheme. The designers of the tree have done quite a good job being consistent with them. A summary can be found here:
I found learning the letters with this mapping to the Latin alphabet very useful, because it makes typing Devanagari on a keyboard with the Hindi Phonetic layout quite easy for the next lessons, e.g., च is obtained with the key 'c'.
"ca in my mind doesn't represent the sound "cha"."
I guess you're not Italian then! Oh wait, I'm German and I would like to see "tsch" written! No, that's not how it works.
so how do you propose to write छ? chha? That would mean a contrast between ch and chha. Very good - bahut achchhā! No, that looks bad. The system employed is to use "h" only to represent aspiration. "c" is not being used for any other sound, so we use it for च। Even just now as I typed it on my keyboard, it was located under "c".
"ca in my mind doesn't represent the sound "cha"."
"I guess you're not Italian then! Oh wait, I'm German and I would like to see "tsch" written! No, that's not how it works."
This is Hindi for English speakers, so it doesn't exacty seem unreasonable to suggest using a transliteration system geared towards English speakers. But really ANY transliteration will seem bad if you try to pronounce it like in English.
The transliteration IS geared towards English speakers as much as possible except where it is illogical to do so.
The OP doesn't seem to understand that transliteration is the act of taking a symbol from one script and rendering as a symbol from another script. It is not a way of transcribing the sounds of a language so that person X can guess at how to make the sounds.
But the OP doesn't want transliteration anyway. They want you to just be thrown into learning Arabic alphabet's idiosyncratic and haphazard was for representing Urdu sounds and you should not use the tool at your disposal, transliteration into Latin alphabet, on your journey to discover the thing. Because it's more "pure" just to learn it directly? Oops, why are we using a computer then. Need to go lie in the lap of your mom speaking Hindustani, rather than get confused by pesky alphabets...
Does your computer display the threads in a "nested" view? If so, you'll see that this post is responding to someone else who brought up transliteration. By the end of my post, after addressing Xwyjz8Ys on transliteration, I write that "But the OP doesn't want transliteration anyway." In other words, saying as you said here that the discussion was not about transliteration, to bring the discussion back to your original post.
I did write "The OP doesn't seem to understand that transliteration is ... " I am sorry for the hasty interpretation if you do understand what transliteration is. The meaning behind my comment is that use of Latin alphabet in teaching Hindi phonology —not the teaching of Hindi language as a whole, but the specific matter of phonology —is there to function as transliteration. Your OP treats it as though the function of the Latin writing is to represent sounds directly (phonetically) and in lieu of other information, whereas it's function is actually to represent classes of sound/distinctions (phonemically) and in reference to (transliteration of) the predominant cue: Devanagari.
I deduced from this that, because you misrepresented how Latin writing was being used that you do not fully understand how Latin script-as-transliteration is used in teaching South Asian languages.
Keep in mind that teachers of South Asian languages as L2 nearly always use transliteration as an auxiliary tool when covering phonology. (Then, they drop it subsequently, as this course does.) These are people who have an actual developed pedagogy, with the insight of other teachers, years of experience, and "big picture" view. It's pretty arrogant to start using the language and telling them they're doing it wrong because you have this seemingly original idea that it would be great to just learn in some kind of purist way. It's possible that the transliteration exists for a purpose that the teacher understands.
Mike made several claims, and I responded to one of them. Their claims:
"sounds in Hindi and Urdu that just don't lend themselves to representation with Latin characters". False. They don't lend themselves to any symbol more than another. In language, a writing system has a convention (learned) relation to sounds it represents. A "voice less palatal stop" looks like this in IPA, i.e. the international system used to represent linguistic sounds: https://en.wikipedia.org/wiki/Palatal_stop Hmm... that seems to look a lot like [c]... It's pretty convenient therefore that the representation scheme in Latin alphabet for that exact sound in Hindi is [c], and whining that one's English language doesn't use that symbol for the same sound is pointless.
"The use of a bar over a Latin character isn't sufficient to mark the many unique vowel sounds." Ooh, unique! How many vowel sounds are there? Surely fewer than in English, so what's the problem? There are only 3 vowels that need to be distinguished by a "bar" in Hindi's Latin rendering, and is obviously is an improvement over not using them. They show short vowel vs. long vowel quite elegantly.
"I think the choice of Latin character is just wrong.[because] (ca in my mind doesn't represent the sound "cha".)" This is the claim I already responded to.
The course uses Devanagari system throughout, and only Latin as an auxiliary while learning the sounds. Given issues with the audio, it makes no sense to exclude the Latin representations that are helping learners along their path to realizing those sounds. Mike has made no argument about why, pedagogically, it is bad to included Latin representation as an auxiliary feature. You are asserting there is no need for transliteration. Do you care to make an argument for why it should be done away with?
Further, I included it in my first reply, by I think you people are conveniently ignoring the assertion Mike made about Urdu. Do you really think people would access (i.e. learn, via this medium) Urdu more effectively without using any Roman character in the process?
I agree. I'm just starting learning the letters and instead of learning which hindi letter represents which sound I'm learning which latin letter represents each sound and then which hindi letter that connects to. Really I'm just adding a step to the learning and I'm sure it's going to bite me in the butt later when I don't know the hindi letters without the latin one.
I don't see the usefulness in learning the transliteration. It doesn't help you learn how to say anything, and I feel it's a distraction from the letter learning, e.g. I noticed that when you have two vowels and one has a line over it, sometimes the hindi characters are the same but there is an extra part to the one which has a line over it in latin characters अ a अा ā This helps me do the duolingo exercices but doesn't help me with the language
Again, nested view; I replied to bourneo. Notice, too, that bourneo used the word/concept "transliteration"; my remark is in that context.
I am asking how bourneo has fared in learning the phonology of Hindi.
In using Hindi, we have to learn Devanagari script, yet unfortunately that helps little with learning phonology.
As always, our number one resource for learning phonology is hearing sound. No one has every disputed that.
However, written communication can be an auxiliary tool, as it confirms what sounds are distinguished meaningfully (phonology). For example, though English has a voiceless velar plosive, the fact that it treats aspirated and unaspirated productions as having no bearing on meaning is suggested by the single symbol /k/. That Hindi treats these two as meaningful is reflected by क and ख when using Devanagari. Yet, the Latinized /k/ and /kh/ —especially in the context of the whole transliteration scheme— provide helpful information about this important phonemic distinctions and in a way that is more systematic and obvious than provided by Devanagari.
So, I would ask bourneo how they are getting along forming an understanding of the Hindi phonemic system without this auxiliary tool.
No problem! Interestingly romanization of Hindi is extremely common in India. Sometimes a billboard in India will be the Hindi language in Roman characters. In this way the the standard has already been set when it comes to the romanization of Devanagari. All language learning tools will use this standard when teaching. The standard definitely has flaws, but It’s a start for any Hindi learner whose mother language uses Roman characters.