It is most likely that the Romani language appeared in the 11th - 12th centuries AD in Eastern Anatolia. The lack of some thoroughgoing studies renders uncertain the exact languages/dialects that participated in the creation of Romani and also whether among them there was the pre-eminence of a certain proto-Romani language. Preliminary studies indicate only that Romani is closer to the Central Indo-Aryan languages.

Did these ancestors use to write? What was their position in the society? The study of the Romani language reveals that the proto-Roma were followers of Dharmic religions, as the basic religious vocabulary is Sanskrit. For example, in Romani, राशाय/rashay (priest) comes from ऋषि/ṛṣi, in Sanskrit, or देवेल/Devel (Divinity) comes from Sanskrit देवा/devā. The Roma respect the purity rules followed by the Hindu high castes, codified in Manusmrti (“The Laws of Manu”), for example those from V.85, 132, 135 regarding the ritual distinction between the upper and lower parts of the human body. The Romani writer Leksa Manush gathered a list of Romani terms that would hint to an allegiance to the Shaivism, like the word त्रुशूल/trushul (cross) which originates from Sanskrit त्रिशूल/trishul, the trident of Shiv.

They were integrated in the organization of the state(s) from those times and also in their broad society. In Romani, थागार/thagar (king) has the Hindi correspondent ठाकुर/ṭhakur (lord, honorific title) and राय/rai (Mister) or रानी/rani (Lady) are words found in both Romani and Hindi. The society from the Northern Subcontinent use abugida-type writing systems from the Brahmic family. These evolved from Brāhmī (earliest known use: 5th or 6th century BC) and later Gupta scripts to contemporary Devanagari, Gurmukhi, Bengali, Gujarati and many others sharing some common characteristics and certain resemblances of the graphemes.

Standardization

Since the Roma left the Subcontinent (11th century AD) they wrote few and with different scripts. Usually, these were the writing systems found in the countries they settled: Latin, Cyrillic, Greek, Arabic. Their use might be considered sporadic, and no firm tradition for writing Romani in any of these alphabets emerged. Romani writers have had a tendency to use whatever script was prevalent among the surrounding people and to use code-switching between more writing systems when necessary to facilitate the communication.

In 1990, a ‘’Commission for the standardization of the language\'’ of the World Romani Congress made an attempt for standardization, upholding the Latin alphabet proposed by Marcel Courthiade. However, since then, it did not gained strong support, the main issues being the exact correspondence between some Romani sounds and the Latin letters that would be suited for them, also about writing the postpositions.

The sounds that have more variants are those that did not exist originally in the Latin alphabet, later being employed diacritics on older letters or simply groups of letters to write them. For example, there are known three ways to write four sounds from Romani:

The first style of writing, with caron, is used in Slovakia, Czech Republic, Hungary, Yugoslavia (maybe other countries, usually Slavic) 3 4 5. The second, with acute accent (the variant proposed by Marcel Courthiade), is used in Romania 6. The third is used in Hungary 7 8 9 and also in all the other countries from Central Europe enumerated above. This seems to be the most used by the Roma in other parts of Europe 10 11 and in other continents, 12 13. Also it is the most used on Internet by users, because diacritics are not easy to find on-line. Besides them, there are also other phonemes with more Latin variants.
Since Romani is an Indo-Aryan language, it use postpositions attached to the nouns in oblique case. The alphabet proposed in 1990 employes different graphemes when the sound is the first consonant from a postposition. These graphemes would cover the difference between, usually, voiceless (at singular) and voiced (at plural) consonants. For example: avreske- आव्रेस्के (to the other) and avrenge- आव्रेन्गे (to the others) would be written avresqe and avrenqe. There was also the hope to cover some minor dialectal differences of pronunciation (in this case: avresce- आव्रेस्चे and avrenje- आव्रेन्जे). The other graphemes are:

- avresθe/θar (at/from, by the other) and avrenθe/θar (at/from, by the others), pronounced avreste/tar-आव्रेस्ते/तार and avrende/dar-आव्रेन्दे/दार;

- avreça (with the other) and avrença (with the others), pronounced avresa-आव्रेसा and avrentsa-आव्रेन्त्सा.
Up to now, they have not gained much popularity, although they are used in the Romani education system in Romania.
Devanagari

The modern use of Devanagari appeared when Romani individuals began to study the Romani language and culture in the Indo-Aryan context, either as self-educated persons or in the higher education available in Romani (mostly in central and southeastern Europe), with a contemporary utilization less than of the Latin alphabet. Devanagari has a correspondence one sound - one letter and its graphemes are suited for the Indo-Aryan sound system (the aspirated phonemes, for example, have distinct letters or the presence of the sound noted by व्, something between v and u in the Latin alphabet). It also possesses a simple and organized structure based on phonology. The evolution of the language from Sanskrit to modern Indo-Aryan languages brought into use few new phonemes as the script evolved from Brahmi to Devanagari. Then, in the last centuries, since Devanagari gained a stable form, there has been only one diacritic for new sounds, nukta applied to a letter marking a close sound. Therefore there is no doubt about how to write the Romani language in Devanagari.

Devanagari is, likewise, the vehicle of a cultural area similar to that of the Romani people, giving important tools for creativity. It does not yet have a tradition of use, the same as do the other writing sytems employed for Romani.
Table of correspondences

Below is the table of correspondences between Devanāgarī and Latin writing systems (there might be also other Latin variants in use):