Ranjana Ta

Ta opens the fourth row of the consonant grid, the dental class, and is its plain voiceless member. A dental consonant is made with the tongue against the back of the upper teeth, further forward than the retroflex row and further back than the labials that follow, and the sound is not quite the t of English, which is articulated at the ridge behind the teeth and falls between the two Indian categories. That is why transliterating between the systems is imprecise in both directions. Sanskrit keeps the dental and retroflex rows entirely separate and builds words on the contrast, so a script for the language must carry both, and the varnamala's decision to place them adjacently makes the pairing easy to learn and easy to check.

Ranjana writes Ta as any consonant sign, with the inherent short a and the head stroke, and the letter is among the commonest in the script because dental consonants are common in Sanskrit and in the vocabulary of Buddhist scripture in particular. The letter is also a good illustration of why hearing a distinction and writing one are different problems. A speaker whose language has a single t hears both Indian rows as that sound, and no amount of attention to the letters will produce the difference until the ear has been trained. The chart can only say which two things are being distinguished.

Writing a Ranjana letter in Latin script is a compromise before it begins, and the dental row is where the compromise is most visible. Scholarly romanisation marks the retroflex series with a dot beneath the letter and leaves the dental unmarked, so t stands for the dental and a dotted t for the retroflex, which is precise but requires characters that ordinary keyboards do not have. Simplified systems drop the dots, which collapses ten letters into five and makes two different words look identical. Doubling the letter, as this atlas does in naming the retroflex series, is a third compromise that keeps the distinction visible without needing special characters.

None of the three is the script; all are attempts to point at it. That is worth remembering for a script in Ranjana's position, because a reader with no Unicode text to consult is dependent on romanisation to a degree that a reader of an encoded script is not. The letters recorded here therefore carry both a transliterated name and a drawn outline, so that the name can be searched and the shape can be checked. There is a related point about names. The letter names used in this atlas are transliterations chosen for searchability rather than the forms a Newar teacher would use aloud, and they are a convenience for a reader working in Latin script rather than a claim about how the script names itself.