A graphemic addition to a written symbol used to create a new symbol from a pre-existing symbol. Economically, diacritics help keep the inventory of basic phonetic signs as small and as comprehensive as possible: for example, in German the diaeresis is used to distinguish between ä, ö, ü for [ε], [ø], and [y] vs a, o, u for [a], [o], and [u]. In the IPA (phonetic transcription), a little circle set below or above a letter distinguishes between voiceless and voiced consonants (e.g. voiceless /b/, /d/, /g/ as , , vs voiced /b/, /d/, /g/ as [b], [d], [g]). In syllabic writing, where there are basic signs with standardized voicing, diacritics can be used to indicate the rest of the voicings (e.g. o in Siamese, a in Hindu writings).
Here are some examples with the Roman alphabet as the basis for new symbols: ā for [a:] in Latvian; ă for in Rumanian;å for [o] in Swedish;á for [aυ] in Icelandic;ñ, Ñ for in Spanish,ø for [ø] in Norwegian;è for [ε] in French; and o for [ŋ] or , respectively, in Igbo. Up to 1976, modern Greek writing was oriented towards ancient Greek such that there were numerous (and virtually superfluous) diacritics. There are also various diacritics in Hebrew as well as in the different orthographies of the Semitic languages. In Indonesian a superscript 2 can indicate reduplication:orang2 for orang-orang (‘persons’) vs orang (‘person’). Diacritics are also used to indicate that the symbol refers to a number as opposed to a sound, e.g. Grk ε’ for 5 vs ε for /e/. (alsoacute accent, cedilla, circumflex, diaeresis, grave accent, tilde)
, 
, 
vs voiced /b/, /d/, /g/ as [b], [d], [g]). In syllabic 
in 
in 
and o for [ŋ] or 
, respectively, in 
