Transliteration and Transcription
When studying foreign languages and scripts, it is important to distinguish between transliteration and transcription. Both can be used to encode language in a familiar script; however, they differ in that:
transcription inscribes uttered phonemes [perhaps] for future rearticulation of the utterances;
transliteration transforms inscriptions between orthographies [perhaps] by rules that preserve reversibility.
Whereas transcription preserves speech, transliteration preserves writings. Phonetic transcription is a phonetically more precise (with respect to the source language) and unambiguous method of transcription.
Usage and examples
The author's choice of method for transliteration or transcription of a (foreign) language often depends upon his perception of his readers' ability to decipher and possibly rearticulate what he will have written.
For instance, in modern Japanese, English loan words are transcribed in Katakana (a syllabary reserved for flashy signs and foreign words), and then interlarded within regular Hiragana and Kanji scripts. However, owing to the paucity of the Japanese language's phonetic inventory, many common English sounds do not have Katakana equivalents. Long loans words are often shortened and most sounds are euphonically approximated in transcription: レッジー シャシュル (rejjī shashiru) for Shashir Reddy; オケ (oke) for orchestra; アニメ (anime) for animation; テレビ (terebi) for television; etc.
In reverse, Japanese is transliterated into Rōmaji (ローマじ or Romanization) somewhat regularly from Kana (Hiragana and Katakana syllabaries). Long vowels are usually but not always marked by a macron (except for the long "o" in native Japanese words — this is usually rendered as "ou", reflecting Hiragana spelling for possible vowel backness). The general nasal consonant (Hiragana ん, Katakana ン, Rōmaji "n") can sound like a velar, bilabial, or (especially before a vowel) a dental nasal, and occasionally a nasalized vowel, but in Kana it is written distinctly from the five strictly dental nasal syllables. This is reflected in Rōmaji, by separating the general nasal consonant and vowels by apostrophe. Note that vowels in Rōmaji are more aligned with Continental European pronunciation (especially Spanish) than English. Consider the following examples: sen'ou/sen'ō (せんおう) meaning tyranny; karate (からて) meaning empty-handed; senpai/sempai (せんぱい) meaning elder/senior; etc.
Since the Kana syllabaries align very well with actually pronunciation of Japanese (as is the case with most non-English and non-French languages and scripts), Rōmaji closely corresponds to spoken Japanese. This correspondence qualifies Rōmaji as a (Latin) transcription system of Japanese and even a phonetic transcription system.
Of course, even Rōmaji breaks down when considering other source languages. Currently, the most universal phonetic transcription system is the International Phonetic Alphabet, which is itself a Latin-based system and near-perfectly unambiguous.