Amitabha and Eight Great Bodhisattvas (Amita Gujon)
1300s hanging scroll
Photo: Wikipedia Commons
Chuang Tzu is the traditional name, but Library of Congress and current usage is
Malibu Hindu Temple, Calabasas
photo: Creative Commons via Wikipedia Commons
Transliterating between two alphabetic languages is one thing. Going from languages written in ideograms to an alphabetic romanization is quite another. So there have been a number of attempts to create systems for writing Chinese, Japanese, and Korean with Latin/Western European letters, within a number of European languages.
For Chinese-to-English, the Wade-Giles system was the one used in your great-grandparents' and grandparents' days. Of course, Chinese is not just one language, but a complexity of language systems going back thousands of years. Cantonese and Mandarin are two major variant dialects of the Chinese language in the way they are pronounced, and there are numerous additional dialects. However, all readers understand the same written/printed Chinese characters because they represent ideas, not sounds.
Before the larger conformity to Mandarin enforced by the People's Republic, a unified system was a very complex thing to attempt. The Yale system was an improvement over Wade-Giles, and had about a 20 year heyday. In 1982, the Hanyu Pinyin (or just Pinyin) system was officially adopted as the system for converting Mandarin Chinese (and the loan words and dialects woven into it) into Roman letters. Peking became Beijing, Mao Tse-Tung became Mao Zedong, and X changed its pronunciation to 'sh'. Here are a couple of Web sites with examples:
Library of Congress description of some differences between systems: http://www.loc.gov/catdir/pinyin/difference.html
Library of Congress: Correspondence of Wade-Giles to Pinyin http://www.loc.gov/catdir/pinyin/romcover.html#7
Sometimes, you just have to use what you find, as different authors and systems just romanize differently. For example, romanizations for a word in the dialect of the ethnic minority group in China commonly known as Uygur or Uyghur… and Uighur in the LC subject heading! ... you can find the variations Wei wu er / Wei’wu’er / Weiwuer in Pinyin. (Thanks to Rachel Wen-Paloutzian for this example.)
For added complexity, the meaning of spoken words in Chinese is conveyed by pitch as well as the sounds used, which is not reflected in romanized transcriptions. This means you could read a romanized transcription clearly, but with the wrong pitch (notes), and wind up in either incomprehensibility or deep trouble. Of course, things could be worse for you: imagine the process the other way around, which is an art form of selecting not just similar sounds but also individual Chinese characters that convey the meaning or descripton of the name or idea being transmogrified!
Japanese romanization is a bit easier because of the way the language is constructed: Japanese hiragana and katakana letters represent sounds (syllables, not freestanding consonants) as well as ideas, and the Japanese use many Chinese characters (kanji) as a kind of shorthand (not, of course, pronounced in Chinese). The runs of syllables are romanized most often using the rōmaji system, though the Hepburn and other systems have also been used.
Korean writing has its own hangul and hanja characters (which go back a very long way), and there have been several romanization systems employed, including one by the Library of Congress that differs a lot from its MR predecessor, and from the Yale entry system. The South Korean government chose the Revised Romanization/Ministry of Culture system back in 2000 as its official system.
Then there are Cambodian, Burmese, Tibetan, Hmong and a host of others. (Vietnamese was romanized during the colonial period.)
The other major language group is what some call Hindustani, but is made up of all the languages and dialects of the massive Asian Subcontinent. Suffice it to say that, within each of the now officially separate languages, Hindi and Urdu (both having two standard variants plus multiple variations), there have been multiple ways to romanize them for a long time. The major written alphabet is devanāgarī, which is used to write most of the languages (with the exception of Gujarati). For romanization, the nation of India has officially adopted the Hunterian system. There is also the International Alphabet of Sanskrit Transliteration (IAST), which is used widely for the transliteration of ancient Sanskrit and Pāḷi. But, as always, there are numerous complications and variations I won't dare go into, as well as additional systems like ISO 15919 and the somewhat simpler Harvard-Kyoto system. (Our Yoga Studies graduate students deal with some of these complications all the time, which garners my respect.) And so, you will see, for example, Siva or Shiva, Visnu or Vishnu, Krisna or Krishna or Krushna, sometimes with a dot under the S to indicate the 'sh" as in Kṛṣṇa (IAST romanization), sometimes without, but sometimes with an accent mark on top, as in Śiva. And then there are words we are familiar with, like nirvāṇa which have close equivalents in a different dialect/language (e.g. nibbāna) which might be the preferred word in some religious texts.
Finally, overall, the problem for searching names in any of these languages is that romanization is not always consistent, even sometimes within the same system, or even in works by the same writer.
LINUS uses the Library of Congress headings as its search vocabulary. Librarian-catalogers spend a lot of time and energy creating cross-references within the catalog to make your searching easier. Try using the 'official' romanized form of the name you find in LINUS or LINK+ as your first choice in searching other databases, though some of the diacritical marks, hyphens, underscores, etc. may have to be altered in some of them. And, when doing an exhaustive search, try using the variations you find in different databases and by different authors, combining them with the OR operator, to open up more possibilities.