Creating script based on Devanagari - PhMota/letters GitHub Wiki

let us try to create a Universal Phonetic Alphabet based on the Devanagari script. The reasoning behing was to select shapes that reflect the tongue, lips and pharinx position and shape during the pronounciation of the phonemes in order to creat a featural alphabet which is easy to learn. The only existing featural alphabet is the Hangul alphabet which combines syllables in blocks. However, Hangul is quite poor in phonemes lacking labiodentals, retroflexes, palatals, uvulars and pharyngeals. Also the approximant, lateral and trill manners are missing. In the vowel side, there are 6 base vowels only. Furthermore, the block forming rule only allows for CVC syllables and a few specific extensions to it which are not generalizable. Therefore, the pleasant simplicity of Hangul cannot be extended to include the IPA.

Instead, the Devanagari abugida surges as an interest contender, since it incorporates base consonant shapes for almost all consonants of the IPA. Only trills, approximants and lateral fricatives are missing. On the other hand base forms are reused and combined in Devanagari to form other consonants. The major and most obvious feature of Devanagari is the horizontal line that runs along the top the letters and connects them into words, the shirorekhā. The shirorekhā will be used loosely to represent the alveolum, palate and velum. For lingual consonants, the base shape will represent the tongue position. We chose to represent the mouth as a sagittal crossesection with the lips to the left and the pharynx to the right. The script is to be scritten from left to right.

The articulation of the consonants is taken as the base for the featural shape choices. In order, we chose

bilabial: u, उ, flipped representing both lips,
labiodental: u, उ, flipped and modified with the connection to the shirorekhā straightened to represent the teeth,
alveolar: marwari dda, ॸ, flipped with the curve representing the tip of the tongue approximating the alveolum. To distinguish dental-alveolars and post-alveolars, the placement of the conection between the curve and the shirorekhā is shifted forwards and backwards, keeping it in the middle for alvealars,
retroflex: va, व, representing the curved tip of the tongue
palatal: dental da, द, flipped representing the middle of the tongue approximating the palate
velar: dental ta, त, flipped and modified removing the lower part of the vertical line, representing the back of the tongue approaching the velum,
uvular: dental ta, त, flipped keeping the vertical line to distinguish from the velars,
pharyngeal: retroflex tha, ठ, flipped representing the lack of tongue with a circle,
glotal: retroflex dha, ढ, flipped representing the epiglotis as the loop,

the manners of articulation are

plosive: long a, ा, is attached after the base consonant; ो for voiceless and ौ for aspirated
nasal: anusvara, ं, below the plosive sign
trill: long i,ी, doubled representing the repetitive touches of the tongue
tap: long i,ी, representing a touch of the tongue
fricative: no addition
lateral: long e, ि, representing the sound passing around the tongue
approximant: drops the conection between the base shape of the consonant and the shirorekhā representing that there is no touch of the tongue

vowels:

rounded have similar shapes to the unrounded but have loops
back vowels are reversed front vowels
there is no shirorekhā above the vowels to represent that the tongue does not touch

	bilabial
plosive

plosive

plosive

These first options have some obvious drawbacks. First of all, most devanagari letters needed to be reversed to accomodate for the mouth to the left representation chosen before. So we will invert for a mouth to the right ─ which is reversed from Hangul ─ in order to keep the shapes of the orginal Devanagari consonats. This may bring some confusion to the native readers of Devanagari, since the shape will be recognizable but associated with a diferent phoneme.

Another drawback is that the semivowels do not resemble the corresponding vowels. In order to solve that, we will associate the front vowels to palatal and back vowels to the velar point of articulation. This effectively transforms vowels in a manner of articulation. There still remains the matter of the openness of the vowels.

Finally, using the place of articulation as the most complex shape of the consonant resulted in very similar shapes being concatenated as in "strengths" were s, t, r, th are all alveolars. So we decided to use the manner of articulation as the base of the letter and the place of articulation as a diacritic.

		B	LD	DA	A	PA	R	P	V	U	Ph	G
		꣣	꣢	DA	ऻ	PA	ी		ॆ	U	Ph	G
plosive	प	प꣣	प꣢		पा		पी	प	पॆ	पॅ
unvoiced	म	म꣣	म꣢		मा		मी	म	मॆ	मॅ		मि
nasal	प्	प़꣣	प़꣣		प़ा		प़ी	प़	प़ॆ	प़ॅ
trill	ल	ल꣣	ल꣣		ला		ली			लॅ
tap	र	र	र		रा		री
fric	त	त꣣	त	तऻ	ता	तो	ती	त	तॆ	तॅ	तॎ	ति
unvoiced	न	न꣣	न	नऻ	ना	नो	नी	न	नॆ	नॅ	नॎ	नि
lateral	ह		ह		हा		ही	ह	हॆ
lateral	ड		ड		डा		डी	ड	डॆ
approx	ढ		ढ		ढा		ढ	ढ	ढॆ
closed i								७	७ॆु
round i								७ु	७ॆ
midcls e								१	१ॆ
round e								१	१ॆ
midopn E								५	५ॆ
round E								५	५ॆ
open a								२	२

strengths: नामाढा१पॆ्नऻना

mão: प्꣣२्७ॆ्

Rio de Janeiro: तॅ७ढॆ पातो७ तो२पा्१ःरा७ॆ

3. Simplifications

The alveolars are very common and our previous choice to represent them with a short a has cluttered the script. We propose a further modification to simplify the alveolar glyphs using just diacritics. Then reasoning to keep the palatals as the most simple forms is based on the vowels. It is meant to keep consistency with a simple vowel shape, however we have created the oddity of having velar vowels as modified front vowels which also overloads the script.

		B	LD	DA	A	PA	R	P	V	U	Ph	G
		꣣	꣤	DA	ॅ	े	ी		ॎ	ि	꣡	꣠
plosive	प	प꣣	प꣤		पॅ		पी	प	पॎ	पि
unvoiced	म	म꣣	म꣤		मॅ		मी	म	मॎ	मि		म꣠
nasal	प्	प़꣣	प़꣤		प़ॅ		प़ी	प़	प़ॎ	प़ि
trill	ल	ल꣣	ल꣤		लॅ		ली			लि
tap	र	र	र꣤		रॅ		री
fric	त	त꣣	त꣤	तऺ	तॅ	ते	ती	त	तॎ	ति	त꣡	त꣠
unvoiced	न	न꣣	न꣤	नऺ	नॅ	ने	नी	न	नॎ	नि	न꣡	न꣠
lateral	ह		ह꣤		हॅ		ही	ह	हॎ
lateral	ड		ड꣤		डॅ		डी	ड	डॎ
approx	ढ		ढ꣤		ढॅ		ढ	ढ	ढॎ
closed i								७	७ॎ
round i								७ु	७ॎ
midcls e								१	१ॎ
round e								१	१ॎ
midopn E								५	५ॎ
round E								५	५ॎ
open a								२	२ॎ

strengths: नॅमॅढॅ१प़ॎनऺनॅ

scraunched: नॅमॎढॅ५ॎपॅ्मॅनेपॅ

mão: प्꣣२्७ॎ्

Rio de Janeiro: ति७ः७ॎ पॅते७ ते२पॅ्१ःरॅ७ॎ

New York: पॅ्ढ७ॎ ढ५ॎढॅमॎ