Historians of Islam/Muslims in pre-modern India face the unenviable and complicated matter of coming up with a coherent and intelligible system of transliteration for scholarship that utilizes sources in Arabic, Persian, Urdu, and more recently Brajbhasha. Persianists will be the first to counter with their own complaints with trying to utilize an efficient transliteration system. The early standard for transliteration in Islamic Studies was with Brill's Encyclopaedia of Islam, new edition (now called 2nd edition). It was not all together an unattractive system though it did have some particular quirks (dj for the letter ج). The main problem with the transliteration system was that it took no account for the very different pronunciation of the letters such ث or ذ that were pronounced ‘th’ and ‘dh’ in Arabic and ‘s’ and ‘z’ in Persian and Urdu but were transliterated as th and dh all the same. The use of the letter ḳ for ق does not appear to have caught on with Arabists even. The Library of Congress offered an amiable solution to the problem by proposing three different transliteration systems for Arabic, Persian, and Urdu, making small modifications for Persian and more for Urdu, while maintaining links between the three. While it concedes to the three languages considerable flexibility, the same name can be transliterated in radical different names, like ثناء اللہ is transliterated Thanā’ Allāh if it appears in an Arabic source, Sanā’ Allāh in Persian and S̲anāʼullāh in Urdu! Furthermore certain names, which are commonly known by one spelling convention end up transliterated completely different when looking in source material. For example the six Mughal dynast Aurangzeb becomes Awrangzīb.
The transliterations systems offered by Encyclopaedia Iranica and Urdu specialists are not necessarily that much easier. The two systems that seem to have won out are the modified Encyclopedia of Islam, 2nd ed. (used by the New Cambridge History of Islam) and the Library of Congress, the latter for obvious reasons of aiding researchers. Yet for South Asian pre-modernists the following considerations and troubles continue: how do you maintain a standard coherent system which while paying attention to source material doesn’t fall into the trap of having transliterations like qāz̤ī-i quz̤āt confuse everyone. Furthermore, since many (for example) Mughal texts used Arabic titles as a practice of prestige, does one transliterate them according to guidelines set by Arabic or Persian? Suddenly a lot of Mughal taz̲kirahs (biographical dictionaries) would become tadhkiras.
It is no surprise that recent works on Mughal history have eschewed transliterations systems altogether, a trend that is also being adopted by publishers who want their books (regardless of whether they are on South Asia or not) to be read by undergrads. Unfortunately there is no easy or happy solution if we choose to adopt a transliteration system, but in the meantime we must beg the forgiveness of Sher Shah, the Sur ruler of India in the middle of the 16th century who is now Shīr Shāh according to the Library of Congress Persian system or, even worse, Shayr Shāh, if one wishes to be dogmatic about South Asian Persian pronunciation.